Coastal Engineering 2002

The following sections are included:

• Foreword

• Proceedings Dedication

• Acknowledgments

• Sponsorship

• CONTENTS

A significant proportion of human development has taken place in the coastal zone. This region is rich in natural assets for supporting human life and facilitates communication and trade. In recent years an increasing appreciation of the natural environment and biodiversity of the coastal zone refocused our perceptions of its value. However, the processes controlling the natural evolution of the coastal strip are dynamic and in many cases lead to erosion and exposure to inundation by waves and tides. Where these have conflicted with development, attempts to stabilise and control the coast by hard engineering means have sometimes led to less than optimal results.

With the present drive to achieve sustainability, and in the knowledge that climate change will exacerbate the present situation, it has become necessary to review our approach to protection of all aspects of the coastal zone to achieve balanced and enduring solutions. This is possible due to our continuously improving knowledge, through scientific research, of the processes in the coastal environment, and our determination to manage the coastline for the benefit of as wide a range of interests as possible. This can be achieved by a cascading approach to planning, from global through regional to local, taking place over ever decreasing intervals of space and time. In this way coherence is achieved between strategic plans and local interventions, avoiding the effects of piecemeal defences seen in the past.

This paper makes the case for strategic planning, at all levels, of the coastline, incorporating all the uncertainties involved therein, and shows, by case study, how such an approach has been applied in the UK. A description of planning at a national level is presented, and then it is shown how such planning can be driven down to a more local application, with the application of science facilitating the decision making at each level.

This paper discusses technical and economical procedures associated with long-range planning of beach nourishment projects. It establishes that under certain conditions, including limited availability of offshore sources, the interval between operations may be reduced considerably by the use of shallow water hopper dredges to the benefit of stability and economy.

An examination has been made of multi-decadal and longer-term trends in hurricane frequency and intensity in the North Atlantic and Caribbean Basins. The investigation points to an increasing occurrence of hurricanes, which may also be more intense. There appears to be a strong link between global warming impacts and these findings. This paper presents a recommended methodology for incorporating these anticipated changes into the development of a design wave climate for coastal infrastructure in the Caribbean.

In this paper we present a methodology for the evaluation of significant wave heights, significant wave periods, maximum winds and minimum pressures, using a modification of the original Hydromet-Rankin Vortex Model, Bretchneider (1990) and Holland (1980). The flooding levels caused by hurricanes on the Mexican coastline are calculated using a two-dimensional, vertically averaged finite difference model to evaluate the storm surge, Bautista et al. (2002) and Hubbert and McInnes (1999). The cyclone model is compared to the data series of 29 cyclones recorded by buoys of National Data Buoy Center-NOAA. The flooding model is compared with observed data from Cancun, Mexico. Both models give very good results. For the analysis, the database consists of 53 years of records, covering 1280 hurricanes; 739 on the Pacific and 541 on the Atlantic. For the extreme analyses of wind and wave heights on the Mexican coasts, maps of the scale and location parameters used in the Gumbel cumulative distribution function and numerical results for a 100 year return period are provided.

The El Niño global climate change is shown to dominate the climate of extreme waves reaching Southern California and to strongly influence that of the rest of the West Coast. The increase in intensity and duration of the El Niños during the previous two decades has resulted in significant increases in annual maximum wave heights and in the number of wave events classified as extreme. A mechanism is identified for steering El Niño storm tracks that greatly influences the relative intensity of impacts on the northern and southern regions of the coastline.

Hourly-waves over a period of 51 years from 1948 to 1998 are hindcast at almost all measurement stations deployed on the coastal areas of Japan using a shallow water wave model based on a backward ray tracing method under the input wind conditions extracted from the NCEP/NCAR reanalysis surface wind data archive. Comparison between 2-3 hourly measurements and hindcasts reveals that the system yields reasonable estimates not only for time series of waves and wave climate parameters over a whole measurement period but also for their annual rates of increase or decrease. The important finding is that the 41-year hindcast waves from 1958 to 1998 indicate hardly any statistically significant trend of wave climate on the coastal areas of Japan, whereas the 51-year waves gives a significant increasing trend probably due to the poor quality of wind data in the first 10 years from 1948 to 1957.

Spatial growth of mechanically generated water waves under the action of stratified wind has been measured in a wind-wave tunnel. The dimensionless growth rate β/f obtained as a function of u_*/C, where C is the phase speed of a regular wave with frequency f and u_* the wind friction velocity, seemed almost the same among unstable, neutral and stable conditions. However, the growth rate β was found to be larger at unstable than at stable condition. On the other hand from the theoretical calculation Voorrips et al (1995) obtained the opposite results, and this may be because they calculated β as a function of u_*/C. In this connection the effects of air-water temperature difference on wind-waves and wind drag are discussed.

Assimilation of satellite observations in wave modelling is briefly described. The results obtained from numerical tests to highlight the impact of assimilation of Synthetic Aperture Radar (SAR) observation are presented and discussed. Two enhancements to the wave modelling physics (namely: wind gustiness and quasi-realistic variable air density) are introduced into wave model. The impact of this change is presented and discussed. The tests of SAR assimilation and enhanced physics show slight but positive impact. Therefore, ECMWF operational model included the enhanced physics described here since April 9, 2002, while the SAR assimilation will be operational soon.

Records of large waves may be missing from data sets. This paper deals with the effects which these missing records have when estimating extreme wave heights associated with specified return periods. The study involves numerical experiments, using Monte-Carlo simulation, in which true return wave heights are compared with estimated values. Effects are examined of changing (i) the parent probability distribution from which the data are generated, and (ii) the ratio of missing data to total data. The importance of knowing not only the type of parent probability distribution but also the value of its shape parameter when fitting to data are also demonstrated.

The use of oceanic radar to measure currents at the sea's surface has become practicable. Many observations of currents in coastal zones have, in fact, provided quite worthwhile results. In contrast to this, the measurement of ocean waves by oceanic radar is still at the theoretical level. Bayesian estimation may be applied to obtain directional wave spectra from Doppler spectra. This method is, in theory, applicable, but noise contaminates the measured Doppler spectra so that they deviate significantly from the ideal. Bayesian estimation relies solely on second-order backscattering components, yet accurate estimation of directional wave spectra from observed Doppler spectra requires the proper separation of the first- and second-order backscattering components of the Doppler spectra; this is a difficult task because of the great variety of observed Doppler spectra. A method where we apply wavelet multiple-resolution analysis to separate out the first- and second-order backscattering components of the spectra is discussed. In addition, a new method where we apply both the first- and second-order backscattering components to obtain directional wave spectra that have better resolution is described. Finally, HF oceanic-radar facilities along the Kashima coast were used to conduct a series of field measurements of Doppler spectra, from which the directional wave spectra were then estimated. Simultaneous observations were made with a wave-gage array, and subsequent comparison showed good agreement with the results obtained from the HF oceanic radar observations.

A modern directional wave measurement network is operating since 1989 around the Italian coasts. The remote-controlled buoy system is managed by the National Hydrological and Marine Survey with excellent results in terms of data acquisition rates, temporal coverage and reliability.

Recently this permanent system has been upgraded and expanded (from 8 to 14 stations) after a careful revision of both the sensors and the software for data transmission, acquisition and analysis. The developments of the monitoring system are now calling for real-time data control and distribution.

The present paper aims to inform about the results of the complete data analysis effort promoted by National Hydrological and Marine Survey and carried out by the University of RomaTre in order to better know the characteristics of waves and related storms in the Mediterranean sea.

The need for more organised wave recording arrangements for the UK has been identified over recent years. In 2001 DEFRA commissioned a Feasibility Study to review the situation and make proposals for setting up a near shore wave network. The Feasibility Study concluded that there was likely to be a good economic case for establishing a long-term network. A business case for a national network was submitted and approved, enabling the establishment of a network to proceed with government funding. The business case indicated that a network of new wave rider buoys to complement continuing data from some existing Met Office buoys combined with HF radar wave measurement installations would be most effective. Detailed design studies were undertaken in 2001/2 to identify technical and administrative proposals for the network and user requirements. Work began in September 2002 to start installing a new network of 5 wave recording buoys and to develop a database for collation and dissemination of wave data. This is expected to be operational by April 2003. Technical studies are progressing into the use of HF radar with the intention to establish a pilot installation in 2003.

This paper describes an improvement in the method based on Local Approximation using Simulated Annealing with a Stokes II wave model (LASA-II), reported by Medina (2001) for decomposing incident and reflected wave trains in wave records. The new add-in to the LASA-II consists in a fitted Stokes V wave model upgrading LASA-II to LASA-V. This improvement makes possible its application to non-stationary and highly non-linear wave fields. This time-domain method has been applied to analyze wave grouping characteristics of the incident wave trains corresponding to the tests conducted in the Laboratory of Ports and Coasts (LPC) at the Universidad Politécnica de Valencia (UPV) on a 1:50 scale model of a new breakwater for the Port of Gijón (Spain). The results were compared with a well-accepted 3-point least squares method for separating incident and reflected waves (Baquerizo et al., 1997). Further, pressure gauge records measured on the crown wall were analyzed and compared to the incident wave height obtained by the two methods.

A numerical coupling between mesoscale meteorological and wave model is developed to predict wave field. Sensitivity to the atmospheric boundary layer of meteorological model on the wave prediction is discussed. The numerical results are compared to field data of significant wave height and bottom period. The mesoscale meteorological model with the turbulence models for atmospheric boundary layer gave good agreement with observational data in comparison with analyzed ECMWF data.

A statistical model for the long-term distribution of wave height duration is developed in this work. The relationship between directional sea state intensity and sea state duration is obtained by means of a mathematical model that describes the bivariate distribution of significant wave height and the duration of the persistence. The application of the methodology to a particular site on the northern coast of Spain shows that the proposed model can be used to determine the wave climate, improving the results from existing formulations.

The paper presents the preliminary results of a research project aimed at creating a coupled wave-current model for the Baltic Sea and the Gdańsk Bay. The coupled model is based on two third-generation spectral wave models WAM and SWAN and the three-dimensional circulation model POM. The models working separately proved successful in simulation of several storm events that occurred over the Baltic in the years 1998-2001. The results of modelling, apart from providing information necessary for the statistical analysis of typical wave and current conditions in the Baltic Sea, serve as a reference data for subsequent coupled models runs, that gives a clue as to how important wave-current interactions are in the areas of interest.

A sequence of analytical solutions explore aspects of response patterns expected from numerical codes for storm tides in one and two dimensional basins. Complete analytical details of the solutions are provided, together with specific suggestions for an associated set of analytical benchmark tests. Illustrations of predicted response patterns provide the basis for a discussion of many significant physical aspects and their representation in discrete numerical codes.

This study proposed a storm surge prediction system in which the relation between wind stresses (wind field), whitecap breaker stresses (wave field) and bottom stresses (current field) is summarized. A storm surge prediction system in the extremely shallow water and/or shoaling region is proposed. The current enhancing effect due to whitecap dissipation of wind waves is considered in the system in terms of the whitecap shearing stress, τ_br, which is an additional sea-surface shear stress for fluid motion caused by wind wave breaking. After the basic numerical experiments, the numerical system was applied to storm surge hind-cast in the Yastushiro Sea, Kumamoto, Japan in 1999, which was a typical case of storm surges in extremely shallow water.

A semi-empirical model is developed for prediction of wave heights pdf in shallow waters. It assumes that a single Weibull distribution holds both in shoaling waters and in the surf zone. Pdf parameters have been linked to water depth and zero-moment of wave spectrum. The model has been calibrated using small scale experiments conducted at the University of Naples and has been validated on data from large scale tests performed at the Grosser WellenKanal (GWK) of Hannover, Germany. A comparison has been done with Composite Weibull Distribution (CWD) proposed by Battjes and Groenendijk (2000). The model seemed to be slightly more accurate in predicting commonly used wave height statistics, except for those with very low exceedence level within the surf zone.

This paper presents the development of a model for the prediction of shoaling wave characteristics up to the point of breaking and their further evolution, including wave set-up, as they break and travel across the surf zone as broken waves. The model is based on the concept of an equivalent linear wave, but non-linear wave heights and selected near-bottom orbital velocity features may be reconstructed from the prediction of the equivalent linear wave characteristics. The beach profile may be plane and of constant slope or it may be barred, and the waves may be periodic or random. Model predictions are compared to and show excellent agreement with experimental observations which were not used in the model's development.

To make clear the relationship between the wave height decay process and the energy dissipation process, internal velocity field after wave breaking was measured using PIV technique, and vorticity and kinetic energy were investigated on the basis of the measured results. It was indicated that the vortices generated at the breaking wave front were separated from the front and were left behind in the wave propagation. Also, those vortices had not only the turbulent kinetic energy but also the ensemble mean kinetic energy. It was made clear that the mean kinetic energy taken out and retained by those vortices largely contributed to the wave height decay process.

This paper discusses the relation between sea surface stress and whitecap dissipation of waves in the shoaling zone based on experimental and observational data. The conducted wind-wave flume experiment shows wave shoaling increases energy transfer rate from wind to waves and currents because of increasing of wave steepness. On the other hand, the observations of flow mechanism under the storm condition (strong wind & high waves) show that vertically uniform longshore currents is developed in the wide area of coastal zone including the surf zone and cross-shore currents with strong shear flow is developed only in the surf zone. From these experimental and observational results, it is found that sea surface stress is emphasized under storm conditions by means of energy transfer from wind to waves through whitecap dissipation of shoaling waves.

It was made clear in this study that coastal currents in the shoaling region under storm conditions should be subject to the sea surface stresses which was determined by two types of energy transfer, from wind to waves, and from whitecap breaking to currents.

We herein introduce an internal and dynamic mechanism instead of an external static mechanism to greatly improve our understanding and application of wave breaking in numerical models. Physical experiments were performed to discover the relationship between the relative trough Froude number and the initiation of wave breaking with and without current conditions. Tests were also recorded in two ways: raw measurement data (water surface elevations and particle velocities) and cinematography using a digital video camera. The recorded data was later analyzed to determine when and where the waves break in conditions both with and without an ambient current. The similarities between the hydraulic jump and bores are used as a basis for the theory. Use of 1.36 as the critical relative trough Froude number at breaking is recommended for numerical models. Additional laboratory experiments are recommended for plain beach slopes other than 1:20 used in these experiments and for bar/trough beach profiles to study how the Froude number criteria performs as water depths increase landward of the bar.

Two spectral equilibrium ranges are identified in the surf zone based on spectra measured in the laboratory and field. For k > 1/depth (where k is wave number), the spectral shape is k^−5/2 (Toba 1973, Kitaigorodskii 1983, Resio et al. 2001) and for k < 1/depth, the spectral shape is k^−4/3 (Zakharov 1999). The measurements cover a broad range of conditions, including plane and barred beaches, multidirectional and unidirectional waves, wave heights from 3 cm to 3 m, and peak wave periods from 1.25 sec to 18 sec. The equilibrium range coefficients for the two ranges are found to be a function of the local wave height-to-water-depth ratio. Laboratory measurements of bimodal wave trains show that complex incident spectra evolve to the same equilibrium range shapes as single-peaked spectra, with the higher frequency spectral peak eliminated in the surf zone through a combination of wave breaking and nonlinear interactions. Wave breaking on a current also exhibits the same equilibrium range shape. The equilibrium ranges have applicability to linear and nonlinear wave modeling and pressure gauge analysis.

A set of fully dispersive evolution equations with exact second-order transfer functions is derived. It is shown how Fast Fourier Transform (FFT) can be used for speeding up the calculation of the nonlinear terms within evolution equtions. A surface roller scheme for wave breaking is incorporated. Comparisons for results of regular breaking waves with a time domain Boussinesq model and a conventional breaking model of evolution equations are presented. The new breaking formulation is somewhat better to reproduce the asymmetric shape of the breaking waves. Amplitude dispersion is investigated using a perturbation approach. It is found that fully dispersive evolution equations overpredict amplitude dispersion severely.

A fully nonlinear 2-dimensional numerical wave flume, based on a boundary integral equation method, has been developed for studying waves at the instant that they break. The flume is used to study monochromatic and multiple frequency energy focussed waves breaking in deep water, and waves breaking due to shoaling over a range of slopes. For each breaking event the geometric and kinematic wave breaking parameters are recorded and examined in detail. It is shown that breaking criteria based on wave steepness are not always applicable, and that criteria based on the particle velocities are only likely to be true for theoretical Stokes' waves. The minimum vertical and maximum horizontal Lagrangian accelerations at breaking are found to be independent of the parameters with which they are compared, with constant values of -g and 1.54g, respectively. It is therefore concluded, that there exists a universally applicable breaking criterion based on these values for the Lagrangian accelerations.

In order to predict the near-bottom kinematics under shoaling and breaking waves, five possible numerical modeling strategies are reviewed and applied to two test-cases performed under regular waves in a wave flume. These tests correspond to spilling and plunging breakers over a plane and smooth slope of 1:20. By comparing the numerical predictions with LDA velocity measurements close to the bottom, it is found that the Modified Transfer Function Method gives good estimations for the horizontal velocity, but it must be emphasized that this approach requires the knowledge of the free-surface signal on input. If only offshore wave conditions are known, best results are obtained from Nonlinear Deterministic Models, in particular from die extended Boussinesq equations. In the breaking zone, however, the agreement with measurement is better for the spilling breaker case; the treatment of plunging breaker needs to be improved in the numerical model.

Phase-resolving, Boussinessq-type wave propagation numerical models require some type of trigger mechanism to initiate (and terminate) wave breaking in the surf zone region. A totally new approach, the relative trough Froude number(RTFN) is presented herein. The RTFN is continually computed using the free surface elevation and the horizontal reference velocity(trough) at each grid point and time. When the RTFN exceeds a critical value, wave breaking begins. And most significantly, when the RTFN drops below the critical value, the wave breaking processes are turned off - automatically. One-dimensional, numerical model results on a plane beach for breaker location are compared with laboratory scale results to confirm a RTFN critical value of about 1.45. This is found also to agree with the theoretical value from the Miche formula.

Recent advances in Boussinesq modeling of nearshore hydrodynamics have offered a platform for the study of wind effects on wave transformation and breaking-generated nearshore circulation. The objective of this study is to quantify the wind influence on coastal waves. The paper documents 1) the parameterization of the momentum flux transferred from the wind to surface gravity waves in the coastal zone, 2) the implementation of the parameterized wind stress into a Boussinesq wave model, 3) the development of empirical breaking criteria with the wind effect, and 4) the test of the extended Boussinesq model against observations. The methodology for the parameterization of the airsea momentum flux as well as the extended Boussinesq model incorporating the wind effect can be used as a tool to study wind effects on nearshore wave propagation and horizontal circulation.

This paper concentrates on observed surface instabilities on the crest of waves breaking over a bar. Experiments were conducted in a 3-D wave basin to determine the correlation between the size of the observed instabilities, herein referred to as “fingers”, and the incident wave parameters. Digital video of the instabilities was captured for each experimental run. Eighteen tests were performed for varying incident wave heights and periods to obtain a reasonable range of incident wave parameters. The digital video was analyzed to determine the size of the instabilities. No apparent dependence of the widths of the fingers on the incident wave parameters was found. The finger widths appear to decrease with time within one breaking wave and do not appear to vary with distance offshore of the crest.

Horizontal and vertical components of velocity have been measured in breaking waves using Laser Doppler Anemometry (LDA) in a laboratory flume. The measurements were made for a regular wave on a plane beach with a slope of 1:20 where the initial wave becomes a plunging and then a spilling breaker. Vertical profiles of turbulence characteristics, including intensities and Reynolds stress, were obtained together with surface elevations at 12 horizontal locations. These data will be used for further analysis of turbulence transport and numerical model evaluation for breaking waves.

A model for solving the two-dimensional Boussinesq type equations including the breaking zone and the swash zone is presented. The enhanced Boussinesq equations by Madsen and Sørensen (1992) are discretised in space using an unstructured finite element technique. The inclusion of wave breaking is based on the surface roller concept and a new procedure for estimating the roller thickness is presented. The simulation of the moving shoreline is based on a technique, where the solid beach is replaced by a permeable beach with a very small porosity. Comparisons with physical experiments for the propagation, shoaling, breaking and runup of regular and irregular waves are presented. The application of the model for large coastal areas is illustrated with a simulation of wave propagation and breaking on a natural beach.

In this paper, the ONR Test Bed for coastal and oceanic wave models is presented, which contains a standarized (fixed) suite of tests, and which can be used as a benchmark and a tool for intercomparison for steady-state and unsteady-state wave models. The test bed cases are taken from welldocumented laboratory and field data and from analytical solutions. In this paper the structure of the test bed is presented as well as the graphical and statistical output which the test bed produces. The steady-state test bed is already available, and the unsteady-state test bed will be released in the Fall of 2002.

Several methods for computing the non-linear energy transfer due to resonant wave-wave interactions are implemented in an experimental version of the SWAN wave model. These methods and their mutual relationships (illustrating their evolution from one to the other) are described. Of these methods, two are addressed in some detail. Of the first, an approximate method called the modified SRIAM method, the accuracy and efficiency are numerically demonstrated for various directional spectra. The second, an exact method called the FD-RIAM, is up-graded from an earlier version (Hashimoto et al., 1998) on the basis of Komatsu and Masuda (2000) to solve an instability problem caused by singularities in the Boltzmann integral. The accuracy and stability of this exact method too are numerically investigated. This FD-RIAM, supplemented with all other processes of generation and dissipation (and triad wave-wave interactions) in SWAN, is applied to the shallow water Lake George in Australia.

To accommodate diffraction in conventional spectral wave models, a phase-decoupled refraction-diffraction approximation is suggested. It is based on the mild-slope equation for refraction-diffraction, ignoring phase information. It does therefore not permit applications in harbours. The third-generation wave model SWAN (Simulating WAves Nearshore) was used for the numerical implementation based on a straightforward finite-difference scheme. Computational results is extreme diffraction-prone situations agree reasonably well with analytical solutions, observations, and conventional refraction-diffraction model solutions. A proper treatment of discontinuities in the wave field (tips of breakwaters) would further improve the results.

Results of a long-term wave measurement campaign in a large, shallow lake (Lake Ussel) in the Netherlands are described. Model validation is one of the main applications of the data. Therefore, the data are extensively compared with results of the SWAN wave model, which is used to evaluate dike design conditions. Wave heights were well predicted by SWAN except in some complex situations. Wave periods were generally underestimated. Finally, there was significant natural scatter in the experimental data. Therefore, model validation should not be based on case studies only. Rather, one should use large data sets for global model validation and case studies for in-detail investigations.

In this paper, we present and test a new wave driver for computations of wavedriven nearshore circulation. Two versions are given, each of which computes unsteady wave refraction, diffraction, and shoaling on strong currents. Extensions are given to include wave breaking and to remove a nonlinear dispersive singularity in shallow water. The two versions differ mainly in their treatment of caustics, with method 2 removing the singularity by means of a perturbation expansion.

Wave transformation is tested by comparing with experimental results over an elliptical shoal. The overall system is tested by computing mean currents in a laboratory rip channel. Results compare well with experiments and confirm significant differences between steady and unsteady forcing. Finally, interactions between wave groups and longshore topographic nonuniformities are shown to have resonances which generate edge waves, providing a possible explanation for edge waves which have been observed in the field.

A convolution method for obtaining the internal kinematic field for waves propagating over a constant depth is derived. The method takes the still water level (SWL) kinematics as input. This input must be provided in the horizontal neighborhood of about three water depths from the water column of interest. The source of SWL kinematics may e.g. be one of the recently developed highly accurate Boussinesq-type formulations. The method is valid for irregular waves of arbitrary nonlinearity at any water depth. Only one horizontal dimension is considered.

A one-way coupling between phase averaging and phase resolving wave models for application in port and coastal engineering studies has been developed. The approach developed provides a more accurate representation of nearshore wave conditions than previously obtained. The approach also improves the efficiency of the modelling effort, by minimising the area represented by the phase resolving model. This paper describes the models, methodology and application to an idealised and actual case study.

A finite depth version of the Discrete Interaction Approximation (DIA) has been developed and implemented in the SWAN model. This modification of the DIA makes the presently used depth-scaling obsolete. The capabilities of the finite depth DIA have been compared with results from an exact technique for the calculation of the nonlinear transfer rate. Firstly, the nonlinear transfer rate was computed for a JONSWAP spectrum in deep and shallow water. Secondly, two growth curves have been computed for a shallow lake with a constant depth of 5 m and 2 m. The results of the computations indicate that for mean kh-values larger than 1.3 no effects are noticeable. Only when kh<1.3 the finite depth DIA yields different results. This leads to small changes in wave period and spreading measures.

In the first part of this work we discuss the accuracy and convergence of velocity formulations for water waves in the framework of Boussinesq theory. We demonstrate that most traditional formulations are limited by a finite convergence radius, while this is not the case for the new class of formulations presented recently by Madsen et al. (2002a-b). In the second part of this work, we extend the new formulations to an uneven bottom and include explicit terms proportional to first and second derivatives of the mildly sloping bottom. The applicability of this extension is verified on the case of partial reflection from a plane slope.

For the development of a new port of the A Coruña Port Authority, an extensive and careful analysis of wave propagation had to be done to determine design wave characteristics along the main breakwater, placed behind an irregular shoal. Preliminary numerical simulations showed that wave focusing takes place on the breakwater, and it was finally suggested to complement the numerical analysis with physical model tests in a multidirectional wave basin. Due to the large dimensions of the prototype area, several numerical models were applied to optimize the physical model lay-out (model scale, boundary conditions, location of wave gauges). All physical model test results were compared with results from a spectral wave propagation model GHOST simulations, showing good agreement on wave amplification in the focusing area behind the shoal. The combination of both numerical and physical modeling improved the knowledge on the problem studied. The observed influence of wave period and directional spreading on wave amplification behind the shoal was similar to the observations made by Vincent and Briggs (1989).

An extensive data set of waves propagating over a shoal on a beach has been obtained from measurements in a three-dimensional physical model. The data set has been used to validate the two-dimensional Boussinesq-type wave model TRITON. In particular the performance of the implemented wave-breaking model has been investigated. Since the unique data set contains a variety of wave regimes, also the modeling of other wave aspects, such as linear dispersion and shoaling, as well as nonlinear wave-wave interaction can be verified.

A new system of gradational breaker index, the value of which decreases as the level of wave height within a wave group is lowered, is introduced to facilitate the computation of random wave transformations in shallow water. An empirical formula is also introduced to account for the energy flux decay in trough areas. The parabolic equation is used to compute wave shoaling, refraction, and diffraction. Computation is made with multiple levels of wave heights under the Rayleigh distribution for directional spectral waves. The new system succeeds in reproducing the random wave breaking diagrams by Goda (1975), and shows good agreements with several experimental results on wave transformations over barred beaches and an elliptical shoal as well as with several field measurements across the surf zone.

Images of sand movement due to Obliquely Descending Eddy (ODE) in the surf zone were taken by using a high-speed digital video camera. Near bottom velocity was simultaneously measured by an Acoustic Doppler Velocimeter. Phase of ODE touchdown, local velocity fluctuation, turbulence intensity and shear stress change were examined. From the sand movement patterns and velocity variation, structure and characteristics of ODE were discussed. It was concluded that the ODE is characterized as a pair of eddies which go down and oppositely rotate along the cross-shore directed axes, and ODE is predominant for the turbulence generation and sediment pick-up in surf zone.

In oscillatory flows there is an instability near flow reversal which was observed in laboratory measurements. In this paper, wave boundary layer instability near flow reversal is described by a sudden increase in the mixing length. We suggest a time-varying formulation for the mixing length which implies that an increase in the wave boundary layer thickness corresponds to an increase in the vertical mixing. We suggest therefore a procedure to compute boundary layer thickness for “n” wave periods. A simple free software was developed.

An approach by which the stochastic properties of the bottom friction under nonlinear random waves plus a relatively weak current can be derived from the irregular nonlinear wave motion outside the rough bottom boundary layer is presented. It is demonstrated how bed friction formulas valid for regular second order Stokes waves can be used to find the cumulative distribution function of individual shear stress maxima for irregular second order Stokes waves. The friction factor for nonlinear random waves is also determined, and an example of the calculation procedure is given.

A new set of quasi-prototype undertow data due to plunging and spilling breakers at a beach bar is obtained. Experience is gained with the use, for the first time, of an ADVP at wave breaking. Dynamics of undertow under spilling and plunging breakers across the bar is observed. In comparison to T_p, the local H_rms is more influent in determining the order of magnitude of undertow. The effect of the surface roller on the undertow is also investigated.

This paper is based on data collected during one fieldwork carried out over two days on a barred beach situated on the southern part of the French Atlantic coastline in presence of a moderate narrowband regular swell. During this field work, significant energy transfers to both higher and lower frequencies have been observed, leading to the generation and the release of harmonics. Analysis of the energy density spectra also reveals the presence of a far-infragravity motion (0.001 < frequencies < 0.01 Hz. Using the “R” parameter, defined as the normalized ratio of total velocity to pressure variance, the relative contributions of the edge+leaky waves, and the shear waves to the total energy has been estimated. Although this parameter allows to clearly show the presence of a propagating wave in the infragravity band (0.001 < frequencies < 0.05 Hz) associated to the difference interactions between the two primary incident waves, the interpretation for the far-infragravity band motion is not so simple.

This paper deals with the appearance properties of large wave height in storm. Several wave data obtained in the Japan Sea are analyzed to investigate the statistical characteristics of wave height. The linear and nonlinear simulation methods of random wave are also performed for the purpose of supplementing the wave data, because the data measured continuously in storm are limited. The energy spectrum of wave, the significant wave height and period are used to simulate the random waves. This simulation method is applied to estimation of the characteristics of large wave heights at other wave observation points.

Water surface oscillations are responsible for processes governing the Surf Zone Dynamics. These oscillations are random variables and therefore, the related processes are stochastic processes. An statistical analysis of magnitudes like celerity and undertow is proposed based on the randomness of the water column components.

This paper addresses nonlinear effects in successive wave crest height statistics in finite water depths. Two models of joint distribution functions of successive wave crest heights are presented. The first model is based on second order Stokes type nonlinearity, while the second model is a parametric model based on second order simulations. Analysis results from previous experiments with changing water depths are included in order to evaluate the models. Comparisons are made between the two models and the experimental results. The nonlinear effects observed in the data are well captured by the models.

The probability distribution of the largest (extreme) wave height in finite water depth attained by a stationary random variable over a period of time is developed. This result is based on order statistics and is derived by using the new probability density function of wave height in finite water depth based on the Maximum Entropy Method presented by Ahn (2000). Analytic expression for the estimation of probable extreme wave height is also given. The newly developed probability distribution of extreme wave heights and probable extreme wave heights are compared with wave data obtained from the Atlantic Ocean Remote Sensing Land Ocean Experiment (ARSLOE) by the Coastal Engineering Research Center, U.S. Army, at its Field Research Facility, located at Duck, North Carolina. Comparisons between the newly developed probable extreme wave heights and the largest wave height obtained from the wave records at various water depth for highly non-Gaussian and nonlinear random waves showed excellent agreement.

We propose a numerical wind-wave tunnel which can simulate a fully developed free-surface turbulent flow induced by wind including wave breaking in order to investigate numerically the turbulence structure and entrained air bubble motions due to wave breaking and to clarify relationship between the turbulence structure and scalar transport mechanism across the air-sea interface accompanying wave breaking.

Large scale Monte-Carlo simulation is conducted to investigate the accuracy of a method compounding the probability distribution and the variance of return value separately estimated by the extreme value analysis for stratified data. The simulation results suggest that a method compounding separate probability distribution gives fairly accurate estimate of return value in cases where the parameters of each distribution are evaluated by use of a model based on the least square method. Another finding is that the Izumiya theoretical formula for compounding separate variance yields more reasonable estimates compared to the Goda empirical formula. Then the availability of a stratified sampling technique in the extreme value analysis for the measurement data of winds is discussed. It appears that the technique does not necessarily improve the accuracy and the efficiency in an estimate of return value, probably due to augmentation of statistical variability associated with stratified sampling for the measurement data of small sample size. But the method yields more efficient estimates of return value in some cases where typhoon-generated winds are dominant compared to those caused by other types of storms.

Environmental and economic sustainability of present and future coastal management schemes on low-lying coasts required a better understanding of the role of waves and wave-driven morphodynamics in mudflat and saltmarsh environments. Previous studies noted a large spatial and temporal variation in wave energy dissipation. To adequately assess and manage the energy dissipation potential of individual marshes, this variability, and in particular the mechanisms behind such variability in the energy dissipation process, need to be better understood. This paper discusses the results of an intensive field study of wave conditions (derived from high-frequency (4Hz) ‘burst’ measurements of water level variations) at twenty-one locations on the marsh/mudflat surface at Dengie, Essex, in the period October 2000 - August 2001. Measurements cover a range of inundation depth and meteorological conditions and allow for the effects of a cliffed/non-cliffed seaward marsh edge; marsh width; inundation depth; vegetation density; and incident wave conditions to be quantitatively assessed for the first time. The results show that increased attenuation occurs up to a distance of several 10s of meters, but little to no further attenuation takes place landward of ca 70m inland of the marsh edge. The results also suggest that previous models of dissipation accurately describe wave height increase over a cliff face but overestimate attenuation over the marsh surface immediately landward of the cliff. Variations in water depth and incident wave conditions are not sufficient to explain the high temporal variability observed in this data set and it is suggested that seasonal changes in vegetation density/structure contribute to sub-annual variability in observed wave attenuation.

The strengthening of existing dykes by using the previous structures as a basis created and will create a high variety of dyke profiles with complex geometries at the German North Sea coast. The determination of their effects on design wave run-up is still a problem both for future constructions and for checking up the design of already strengthened dykes with improved design procedures due to the state of the art. The development of mathematical models for wave run-up and overtopping on dykes allow detailed research on the effect of the variety of the geometry of dykes on wave run-up and overtopping. In order to check the reliability of these tools a research project has been started which is jointly carried out by six cooperating European institutions.

This paper presents a brief description of three finite element numerical models based on different versions of the time dependent nonlinear mild-slope equation, presented by Nadaoka et al. (1994), namely the one-dimensional, unidirectional and the two-dimensional versions of that equation. The Finite Element Method (FEM) is used for spatial discretisation while for the time integration a predictor-corrector finite difference scheme is used. The problem is formulated for the free surface local acceleration or the vertical velocity according to the version used. The resulting linear system of equations is solved by a direct method. The free-surface elevations are computed either by a double or a single integration in time, depending on the order of the time partial derivative. Implemented boundary conditions include prescribed freesurface displacement, wave generation by normal-velocity specification, radiation conditions and generation-radiation condition (applied to open boundaries).

The models are suitable for the propagation of surface waves over gently varying depths, taking into account the combined effects of nonlinear refraction and diffraction of waves, not only for shallow waters but also for deep and intermediate waters.

The models are validated with simple test cases of wave propagation over different depths – varying between deep to shallow waters – and their results are presented and discussed. Further, some numerical techniques for implementing the outgoing boundary conditions are investigated and discussed.

A procedure is presented for the modeling of the propagation of waves induced by a sailing ship in shallow water. The method investigated consists of the coupling of two existing models: a 3D potential flow model to calculate the wave field near the ship; and a 2D Boussinesq-type model for the calculation of the wave propagation away from the ship. The study of this type of problems is important in situations where ship-induced waves may cause damage to revetments or hindrance to moored ships. Depending on the bottom topography and the distance at which the ship passes, the wave height near the location of interest may be substantially larger than in the direct vicinity of the sailing ship. Also the wave pattern may change significantly as a result of the shoaling and refraction over an uneven bottom in combination with nonlinear effects and possible wave breaking, and diffraction and reflection at groins and quay walls. The results of several feasibility and validation studies indicate that by combining the strength of both models through a coupling, a powerful tool is obtained for the prediction of ship-induced wave characteristics.

This paper examines the effect of current-induced breaking on the modeling of wave breaking dissipation near rip currents. Simulations using a set of existing wave transformation models are compared to existing laboratory measurements made on a bar/channel topography in the presence of both strong and weak rip currents. The model/data comparisons demonstrate that existing models can effectively simulate the wave breaking near rips, but only by making large adjustments in the model coefficients used in the wave dissipation parameterizations. Since wave breaking formulations are typically calibrated without regard to the cross-shore currents, these results suggest that most model predictions will suffer large errors near rip currents. Furthermore, any errors in the modeled wave breaking dissipation will affect model predictions of the wave setup/setdown and, therefore, the wave driven circulation. A simple model exercise shows that predictions of the longshore pressure gradient, which is a driving force for the rip circulation cells, can be sensitive to the accurate simulation of wave breaking near rip currents.

The dependence on geometric as well as on climatic parameters in the circulation pattern in a semielliptic bay, using the approach by Baquerizo and Losada (2002) is studied. A circulation, mainly driven by the oblique angle of incidence, consisting in two cells with a an outgoing rip current at the axis of the embayment is obtained. The geometry of the ellipse is found to influence the circulation pattern, decreasing the intensity of the central outgoing rip as the ellipse flattens and reinforcing the longshore current at the lateral zones.

Computations of low-frequency motions with a non-linear shallow water model operating on the time scale of wave groups are compared to measurements obtained during the RIP-current Experiment (RIPEX) in conjunction with the Steep Beach Experiment (SBE) performed in the spring of 2001 at Sand City, Monterey Bay. Comparisons focus on wave and flow parameters representing the intensity of the low frequency motions such as the low-frequency root mean square wave height and velocities. A distinction is made between very low-frequency (VLF) motions, with periods in excess of 250 seconds, and low frequency (LF) motions, with periods between 25 s and 250 seconds. The comparisons show the ability of the model to predict both LF-motions and VLF-motions within the rip-channel throughout a number of tidal cycles.

This paper describes results of field investigation on rip currents obtained in April-May of 2001, as part of a RIP current field Experiment (RIPEX) in concert with a steep beach experiment (STEEP BEACH -Thornton et al. in these proceedings) at Sand City, Monterey Bay, CA. These experiments provided a unique field data set of documenting rip current behavior which was comprised of instruments (pressure and velocity) in two primary arrays. The cross-shore array was composed of primarily pressure sensors on the shore connected shoal to measure cross shore wave motion, while a second array transected both shore connected shoals and rip channels. Experimental results observed two energetic motions; 1) rip current pulsations within the infragravity band >(0.004–0.04 Hz), and 2) very low frequency motions (<0.004 Hz). This paper will complement a modeling effort on rip current pulsations and very low frequency motions by Reniers et al. (in these proceedings).

A coupled wave-hydrodynamic modeling system, comprised of a random wave model driving a quasi-3D nearshore hydrodynamic model, is described. Random wave formulations for several inputs to the hydrodynamic model are developed. An alternate wave dissipation mechanism is incorporated into the random wave model, and two wave roller descriptions are implemented to calculate volume flux and other roller-dependent input properties. Comparison to laboratory and field data indicate that an evolving roller description, in conjunction with the 3D dispersive mixing inherent in the hydrodynamic model, yield the best results. A method to nest the model system inside larger-scale wave models is described, and an application to an area of complex bathymetry shown.

Nearshore wave-induced currents at rhythmic bedforms strongly influence transport and mixing processes. Previous research has identified a clear need for the accurate and reliable prediction of strongly sheared nearshore flows where steep gradients, fronts or even discontinuities can occur in the physical flow variables. This paper presents a finite volume Godunov-type numerical solver of the 2-DH period- and depth-averaged equations conserving mass, momentum, wave energy and kinematic wave number. The finite volume model is based on hierarchical quadtree grids that adapt dynamically according to the magnitude of the velocity gradients. Results for the numerical model are in close agreement with laboratory data for wave-current interaction at a tri-cuspate beach in the United Kingdom Coastal Research Facility (UKCRF). A numerical investigation into the nearshore wave-induced currents demonstrates their sensitivity to changes in incident wave angle, assumed turbulent eddy viscosity and the number of repeating cuspate bedforms. A reduction of Thornton's (1970) eddy viscosity parameter from 1.0 to 0.25 is shown to produce rip-like currents cutting diagonally across the surf zone. It is shown that the three cusps used in the UKCRF is representative of infinitely repeating cusps, provided Thornton's eddy viscosity parameter is sufficiently large.

The generation of large-scale, horizontal eddies in correspondence of a submerged breakwater and their evolution in the region shoreward of the breakwater are studied by numerical and analytical means. Such eddies are thought of significantly influencing both flow patterns and sediment transport in the region of the very shallow waters near to the shoreline. Preliminary results reveal the main stages of possible scenarios for the generation and evolution of eddies at the breakwater location. The main characteristics of the transport of passive tracers and bottom sediments associated with the vortices motion are also illustrated.

Mean water level, pressure and velocity measurements of a small rip current system taken on Moreton Island, Australia in December 2000 are presented. During high tide, gradients in the mean water level over the bar and in the trough are weakly directed away from the channel thereby producing no feeder or rip currents. However as the tide falls, the gradients strengthen toward the channel, generating feeder and rip currents. Numerical modeling of the rip system reproduces similar behavior of the circulation patterns. The model indicates that the changing breaking pattern as the water level decreases with the falling tide is the primary factor for the differing hydraulic gradients and determining whether a rip current is generated.

A review of previous research and experiments with numerical models have been performed to determine the range of influence of a tidal inlet on the adjacent unbroken coast. The analyses demonstrate that the influence on the coast extends well beyond the range of the ebb tidal delta. Two aspects of the influence are investigated. The first concerns the residual flow along the coast. The second relates to the wave sheltering effect of the ebb tidal delta. An outline of a conceptual model is introduced based on a partitioning of the coastal and tidal inlet area.

A curvilinear version of the quasi-3D nearshore circulation model SHORECIRC is used to simulate nearshore circulation at Grays Harbor, Washington. We use the measured nearshore bathymetry of Grays Harbor and generate a curvilinear grid that can resolve the complicated geometry and fit curved shorelines and jetties. Tidal currents are simulated by nesting the nearshore circulation model into a large-domain circulation model ADCIRC. Model/data comparisons of tidal currents are made at the measurement gauges located at the harbor entrance. Good agreement is obtained between the measurement data and the numerical results during most of the simulation period. It is found from the measurement data that, under certain wave conditions, strong residual flows appear at the entrance, which may not be predicted by the numerical model with only tidal forcing. Further tidal current simulations coupled with wave forcing improve the model/data comparison and thus show evidence of the short waveforcing effect at the entrance.

The Gamo Lagoon is located in Miyagi Prefecture, Japan. Aerial photographs have been being taken for the coastal area including the lagoon every month or every two months since 1990 up to the present. These aerial photographs are used to detect the occurrence of wave overtopping and resultant sediment intrusion into the lagoon during storm waves. Furthermore, overtopping rate is estimated considering the height of the sand barrier. In collaboration with surveying data, the volumetric concentration of sediment in the overwash flow is estimated. It is found that the estimated concentration obtained in the present study is comparable to laboratory experiment by Kobayashi et al. (1996) in a small-scale wave flume.

The driving force of three-dimensional nearshore currents can be explained by only the vertical distribution of radiation stress. Based on this principle authors have proposed the three-dimensional nearshore currents model (Nobuoka et al, 1998). The inclination of sea bottom must affect the vertical distribution of radiation stress. However, the authors' model did not include the effect of inclination of sea bottom on the vertical distribution of radiation stress. In this paper, the effect of sloping bottom on the vertical distribution of radiation stresses is examined experimentally and theoretically. Finally, the nearshore current model is improved by taking account of this effect and is applied to the surf zone around a coastal structure.

This paper describes a numerical model developed to efficiently and accurately solve the Euler equations in the 2-D vertical plane. The governing equations are transformed from a spatially and temporally varying physical domain to a fixed computational domain by the use of suitable terrain following coordinates. We first solve the hydrostatic problem by separating out the hydrostatic balance from the vertical momentum equation and neglecting the non-hydrostatic pressure contribution throughout the fluid. High order compact finite differences in the horizontal and a Chebyshev spectral collocation method in the vertical are used to discretize the governing equations numerically. Results are obtained for a simple test case of water sloshing in a rectangular tank and compared qualitatively to experimental observations. The model is extended to include a full representation of the non-hydrostatic pressure field by integrating the vertical momentum equation.

In this paper, we utilize the quasi-3D nearshore circulation model SHORECIRC to simulate finite amplitude shear waves. Our purpose was to investigate how the 3D dispersive mixing and the lateral mixing provided by the shear waves affect each other. To achieve this, 2D and quasi-3D numerical experiments are carried out simultaneously. Our calculations showed that both the shear waves and the 3D current pattern contribute to the momentum mixing, and the momentum mixing provided by the shear waves is sometimes larger than that by the 3D dispersive terms. The kinetic energy balance of the shear waves shows that the 3D dispersive terms will extract kinetic energy from the depth-averaged shear waves. Furthermore, the enstrophy equation demonstrates that the 3D dispersion terms play the role of vortex tilting, which allows three-dimensional vortex interactions.

The wave-induced longshore current has important implications on several engineering-related coastal problems, such as the longshore sediment transport, pollutant dispersion, shoreline retreat, etc. Although several theories have been developed in order to predict the longshore current speed, none of them seems to be exempt of drawbacks and limitations. In this paper the conventional momentum balance is revisited with emphasis on the bottom stress. This allows a “better” assessment of friction coefficients and an alternative approach to evaluate the current velocity. The approach has been validated with numerical results and field data from the Spanish Mediterranean coast.

In the paper results of basic investigations on wave run-up in oblique waves performed in the wave basin of NRC-CHC (National Research Council - Canadian Hydraulic Centre) are presented.

In all tests a smooth 1 on 6 slope of the dike in constant water depth of 0.5 m was used. The wave height was constant H_m0 = 0.1 m, while using 4 different wave periods (T_P = 1.26 s; 1.46 s; 1.78 s; 2.53 s) for the generated TMA spectra. The angle of wave attack θ ranges from 0° to 40°. The directional spreading σ _θ was 0°, 10°, 20° and 30°.

The directional function was determined by non-linear regression of wave data measured at the toe of the dike and was calculated as γ_θ = 0.67 cos(θ) + 0.33 both for long crested and short crested waves. Considering refraction and shoaling on the dike up to the breaking area, the directional function proves to be of the cosine type γ _θ = 1.06 cos(θ) − 0.06 ≈ cos(θ) as well for long crested as for the short crested waves.

Approximate shoreline boundary conditions for wave-averaged models of nearshore hydrodynamics are derived on the basis of the swash zone equations of Brocchini and Peregrine (1996). They allow for explicit computation of a non-zero mean water depth at the mean shoreline. This is computed in terms of the local height of the short waves. Implementation issues are also discussed.

This paper presents a breaking waves model in which the wave boundary layer and the swash zone are also included in the model domain. The model is based on the Reynolds equations for incompressible turbulent flows. The numerical solution of the model is performed in a non-uniform grids system so that high resolution can be obtained in the near-bottom area, where the velocity gradients are very large due to viscous effect. Numerical results of the present model are verified by laboratory data for the surf zone, the bottom boundary layer and the swash zone. The comparisons show that the model can predict reasonably the water motions in the surf zone, the bottom boundary layer and the swash zone.

The swash zone is an area containing high fluid velocities and large sediment concentrations and is therefore highly important for shoreline morphology. This paper concerns a large swash zone field study conducted by the University of Plymouth and Loughborough University, the aims of which are to identify the relative importance of swash zone sediment transport processes under various different conditions. The project is an integrative approach, involving many types of data collection techniques, namely high frequency, nearbed velocity, suspended sediment concentration, pressure and turbulence measurements in addition to video, water table and multi-scale morphological measurements. Preliminary observations at a gently-sloping beach in highenergy, dissipative, infragravity-dominated conditions suggest a possible crossshore sediment transport divergence point, the position of which could be due to the balance between onshore and offshore transport mechanisms. The dominant processes in these conditions could be bore turbulence (onshore) and infragravity backwashes (offshore). Water table measurements show that, in this case, inexfiltration is unlikely to play a major role in the sediment transport.

Water waves overtopping a truncated plane slope are considered. The single swash case considered by Peregrine & Williams (2001) is extended numerically to consider multiple swash events. Wave sets of random amplitudes with different periods are considered. A comparison is made between the two cases of overtopping and no overtopping. It is shown how the swash motion in the two cases is different.

a new set of equations using velocity potential is derived based on nonlinear mild-slope equations, which can compute fully-nonlinear and fullydispersive wave transformatioa The numerical model of the shoreline boundary is developed, so that this set of equations can predict the wave profile unison up to the swash zone with the appropriate breaking dissipation terms. The validity of the model is verified by comparison with analytical solutions and experimental data including two-dimensional cases.

In this study the Particle Image Velocimetry (PIV) technique was used to look at the velocity field in the inner surf zone close to the shoreline, under monochromatic erosional (storm) and accretional (swell) waves, standing long waves and wave groups. Differences in the turbulent structure, vertical distribution of undertow, and the shape of orbital velocities of the storm and swell wave fields and their corresponding effects on transport of sediments were discussed. Swash motion of standing long waves and wave groups were also observed.

Laboratory measurements of surf and swash zone hydrodynamics were recorded in the US Army Corps of Engineer's Large-scale Sediment Transport Facility and compared to 2 numerical models to investigate the model's ability to predict water surface elevation and wave orbital and swash velocities. One numerical model relies on the depth averaged non-linear shallow water equations (ID), while the second is based on the full Navier- Stokes equations and uses a volume of fluid approach in the numerical solution (2D). Both models were able to adequately predict the sea surface elevation across the inner surf zone. In general, there was strong similarity between both models and observed velocities with some velocity over prediction during bore passage. Similarity between the 2 models and only small differences between vertically separated near bed (z = 1, 3 cm) velocity time series suggests that the flow in the swash zone is nearly depth uniform for most of the swash cycle implying a thin boundary layer. The 2D model improves on the ID model by allowing for predictions of this boundary layer structure in addition to a more accurate representation of wave breaking and shear stress calculations based on near bed velocity gradients.

The interaction between swash hydrodynamic asymmetry and sediment transport, as evidenced by morphological change in cross-shore beach shape, is poorly understood. Recent advances in coupled swash zone hydrodynamic / morphodynamic modelling, including the BeachWin model, have the capability to provide greater understanding of the relative importance of infiltration on the development of asymmetric uprush / backwash flows. An important requirement is for accurate measurements of water volumes and velocities to assist in the validation and calibration of such models. This paper describes the development of a digital video measurement process to aid in this process. A system using new digital data transfer protocols (IEEE1394) is prototyped and used to capture water surface measurements from a laboratory beach. The data captured during test experiments for a single swash event on a rigid, impermeable beach are then compared to water surfaces generated by the BeachWin model. The results of this suggest that the video data will help to validate the model over more extensive tests.

The time-dependent cross-shore sediment transport model CBREAK is used to investigate the sediment dynamics on a steep foreshore slope (approximately 1:6) in the swash zone on an equilibrium beach. The computed resutls show that the forcing mechanisms and characteristics of suspended sediment dynamics change drastically from the surf zone to the swash zone at the downrush limit. Contrary to the sediment suspension in the surf zone dominated by intermittent high suspension events under irregular breaking waves, the sediment suspension in the swash zone is dominated by the bottom friction due to the uprush and downrush of individual waves. In addition, the sediment concentration and suspension rate are sensitive to small water depth in the swash zone. The computed results are qualitatively realistic but need to be verified against measurements.

Attempts are made to infer an infiltration rate distribution in the swash zone from measured head level profiles using a modified Boussinesq equation model. Limitations of this particular application of the model are illustrated as inconsistencies near the boundaries of the infiltration zone. The effect of the hydrostatic flow assumption in the derivation of the Boussinesq equation is investigated via numerical experiments with a simple 2D vertical flow model.

A series of field experiments were undertaken in 1998-2001 to investigate the interaction between surface and subsurface flows in the swash zone and bed level changes on the intertidal profile on a sand beach (Canford Cliffs, Dorset), a mixed sand and gravel beach (Seaford, East Sussex) and a gravel beach (Slapton Sands, Devon). Data collected included surface pressures under uprush and backwash, subsurface vertical and horizontal porewater pressures, uprush and backwash velocities, and bed level measurements at 10 minute intervals. This paper reports on the measurements from the gravel beach and presents comparisons of swash hydrodynamics on the mixed sand-gravel beach at Seaford and the gravel beach at Slapton. The bed level changes measured on the gravel beach were larger than those previously reported on sand beaches, with an increase of 0.15 m in three minutes and a maximum total elevation change of 0.38 m over the whole measurement period.

A one-dimensional numerical model of the swash zone has been extended to include flow within a porous layer. The effect of exchange between the two water layers (free-flow and porous-flow) on the bed shear stress has been included in the model by solving the equations of motion for the bottom boundary layer with a modified velocity profile. Applications of the model for design of coastal defence structures are given.

Spring-neap tidal water table fluctuations (SNTWF) are oscillations of a period of approximately 14.79 days. These oscillations propagate much further inland than semi-diurnal and diurnal signals. Previous studies have shown that SNTWF are generated as a result of interactions between semi-diurnal lunar and solar tidal oscillations, induced by the moving seaward boundary condition. This paper investigates how the formation of seepage faces results in a spring-neap forcing oscillation on the boundary, and how the propagation of this signal in the aquifer also leads to SNTWF.

Several magnitudes at three different time scales have been studied using experimental results obtained in a wave flume for breaking waves on a 1/20-sloped bottom covered with gravel-bed. Two gravel sizes have been used to study the role of the stone size and bottom permeability. Several velocity profiles inside and outside the surf zone were obtained using Laser-Doppler-velocimeter (LDV). Two sources of turbulence generation have been observed: one at the free surface due to the breaking process and the other due to the porous bottom. The first one is predominant. Bottom-generated turbulence is associated with the vertical velocity gradient and depends on the gravel size. Turbulent scales are studied applying wavelet analysis to the horizontal velocity time history.

As a first step towards the improvement of the characterization of long wave activity at the southern coast of mainland Portugal, a numerical model was developed to simulate wave propagation perpendicular to the coast. This numerical wave tank solves the non-linear initial boundary value problem associated to wave propagation by means of a direct boundary element method. Some preliminary results of the study on the influence of bottom morphology on bound long wave propagation are presented in this paper, which includes also a characterization of the wave storms at the Algarve. The numerical results together with field measurements, still to be carried out, will enable the assessment of the free long wave energy fraction that is produced by the wave group shoaling.

An artificial bathymetry, corresponding to the longshore bar at the main cross-shore instrument array during the Delilah field experiment, has been used as the basis for performing 3.75 hour simulations of wave-driven nearshore circulation for a range of incident wave and frictional characteristics. Results indicate that the Boussinesq model predicts energetic low frequency gravity wave and shear wave climates which are qualitatively similar to results observed in field data. A comparison of two cases with widely differing directional spread do not indicate any dependence of longshore current or low-frequency energetics on the degree of directional spreading.

This paper describes the measurement and analysis of long period waves to support the design of coastal facilities. The suitability of sub-surface pressure sensors for the measurement of long period waves is discussed. The analysis involves the calculation of significant wave heights in a number of discrete wave period bands from 15 seconds to just over 5 minutes. Wave height are calculated from the variance in band pass filtered time series, and from high resolution spectra. Spectral analysis techniques that are used as standard for the analysis of short period waves are not appropriate for this application. The criteria produced from the analysis are suitable for vessel response simulations, marine facility and coastal infrastructure design.

In this paper, the nonlinear SHORECIRC model is used to simulate the laboratory flume experiments by Boers (1996) of random waves on a barred beach. Cross-correlation plots show that the incoming long waves lag the shortwave envelope by 180 degrees in deep water and that this lag increases as the waves propagate onto the slope. This increasing phase angle allows for work to be done on the long waves by which process the infragravity (IG) wave wave height already increases on the slope outside the surfzone. For the case of the Boers experiment it is shown that the generation of the IG waves in the offshore region (before breaking) is much more important than the generation of IG waves by the breakpoint or in the surfzone. It is shown that the relative importance of offshore to surfzone forcing can be estimated using a single parameter. Finally, the energy equation for the long waves is used as an evolution equation to predict the growth and decay of the incoming and reflected IG waves. The results from the predictive model - given the assumptions of a plane beach and bichromatic forcing - reasonably reproduce the SHORECIRC results for the Boers case.

The origin of seiches in the Port of Rotterdam is investigated with observations, weather maps and numerical simulations. It was found that all seiches in the time interval investigated (1995-2001) coincided with either a sharp or a gradual cold front. It was also found that a ‘coldfront indicator’, based on air temperature and precipitation, can be used to detect the relevant cold-front passages and the corresponding onset of the seiches. This indicator is therefore expected to be a straightforward predictor in operational settings. Numerical simulations with a hydrodynamic model forced by atmospheric pressure and wind fields from a highresolution meteorological model (12 km resolution) reproduced the generation of the seiche-inducing long waves at sea for the sharp cold fronts. Experiments for the gradual cold-front cases are ongoing.

Hazardous surf beats appear on coral reefs in the Ryukyu Islands (southern part of the Japanese islands). Their periods extend from some ten seconds to several minutes, which are far longer than the incident wind wave periods. In this study, roles of the 2nd order bound waves on- and offshore the reef are investigated for the generation of the surf beat applying the green function theory. Free low frequency waves are generated due to the discontinuity between the bound waves systems at the reef edge. When the wavelength of the new low frequency wave satisfies the relation L = 4ℓ / (2n–1), (n = 1,2,3, ⋯, ℓ: width of a reef flat), resonant oscillation takes place on the reef. The new low frequency wave is amplified and attacks beaches on the reef. The present model is verified by the measurement in the last.

This paper presents new laboratory data on long wave forcing by random waves breaking on a plane beach. The (outgoing) surf beat shows a strong frequency dependence and is linearly dependent on the incident short wave amplitude. Seaward propagating long waves are also positively correlated with incident short wave groups. Skewness in the long wave surface motion is negative seaward of the breakpoint, but positive inside the surf zone. The data suggest that a time-varying breakpoint is the dominant long wave forcing mechanism for the experimental conditions considered.

Spatial evolution of nonlinear deep-water wave groups is studied experimentally and numerically. Spatial two-dimensional version of the Zakharov equation describing evolution of deep-water gravity waves is used to derive two 4^th order evolution equations, for the amplitudes of the surface elevation and of the velocity potential. The scaled form of the equations is presented. The experimental results for wave groups with initial narrow spectrum are compared with the computations based on the unidirectional Zakharov equation and the Dysthe model. The very good agreement between the computational results based on both models with the experiments prompted an attempt to perform simulations for a wider initial spectral width, that formally violate the assumptions adopted in the derivation of the Dysthe model. The accuracy of the results based on the Dysthe model is checked against the solutions of the Zakharov equation, which is free of restrictions on the spectral width. Conclusions regarding the domain of validity of the Dysthe model are drawn.

Complicated characteristics of tsunami induced flood at a vertical seawall, including overflowing jet and run-up over a hinterland are reproduced by the gridless Lagrangian flow model, or the particle method. Detailed behavior of flow, which is characterized by an overflowing jet, circulating flow cell and a hydraulic jump, is simulated on both of a uniform slope and a horizontal bed with bar-type roughness.

Numerical simulations with a fine grid are conducted in a case study to analyze the whole processes of the tsunami generation induced by a submarine earthquake, the wave propagation, the wave overtopping from coastal structures, and the wave run-up and flooding in a coastal city. This study investigates in detail the effects of disaster prevention facilities on the tsunami run-up and the behavior of the flood flow in an urban area, and it provides the valuable information for establishing the countermeasures against tsunami-induced flood.

After the Papua New Guinea tsunami event, some researchers have studied the landslide-generated tsunami vigorously. However, useful numerical model for prediction considering granular motion has not been developed yet. In this study, a new numerical simulation model for analyzing liquid-solid flows with rapid and great water deformation is developed. The 3-D discrete element method and the 3-D volume of fluid method are used in the model for analyzing the granular motion and tracking the water surface respectively. After validation of the developed model by comparison with experimental results, some calculations for different initial locations of particles are conducted and the results are presented. Further wave generation process and wave propagation characteristics are discussed.

The propagation of tsunamis in coastal areas and the mechanism of the tsunami runup on land have not been sufficiently clarified until 1970s. In 1980s the design formula of the tsunami wave pressure acting on breakwaters was proposed. In this study, to propose a practical formula for evaluating the tsunami wave force acting on land structures, non-distorted hydraulic model tests were carried out. The authors devised the pump-type tsunami wave making systems. This system makes it possible to carry out non-distorted model experiments, which is necessary to evaluate the tsunami wave force. Through the hydraulic model tests the following findings were obtained. Two kinds of tsunami waves were observed in shallow waters. The one is a wave head of which is split into several short period waves, and the other is the wave without them. The former is called as the wave with fission and the other is called as the wave without fission. It was found out that existence of fission gives significant effects on wave pressure. When the tsunami wave is wave without fission, the vertical distribution of the maximum wave pressure can be expressed by a linear relationship. When the tsunami wave is wave with fission, the distribution can be expressed by bilinear relationship. The parameters used for the relationship for wave pressure were evaluated based on the characteristics of the runup water surface elevation without any land structures.

The distribution of pressure and wind in a real typhoon is investigated, using the data from four typhoons, attacked Kyushu Island, Japan. The validity of Myers radial symmetric pressure distribution is examined. An attempt is made to parameterize the observed distortion in the pressure field of the real typhoons. Analytical expression for the modified Myers pressure distribution is proposed. The influence of the distortion of the pressure distribution on the wind field in the typhoon is investigated. The response of the storm surge model to the corrected pressure and wind fields is examined and compared with results of previous model. More accurate estimation of the storm surge is proposed in this work by incorporation of the transformation of the typhoon structure due to land approach and landfall.

In the present design of coastal defense facilities in Japan, the combination of the maximum storm surge and significant wave height by a particularly violent typhoon is often adopted as the design condition. The safety degree is evaluated by the safety factor or the wave-overtopping rate for the design condition. However, the maximum storm surge and significant wave height do not necessarily appear simultaneously. The magnitude of coastal defense failures is related to the duration of the storm. To solve these problems, the storm surges and waves in Tokyo Bay, caused by the major typhoons over the past four decades and the model ones based on the Isewan Typhoon, are simulated by numerical model. The result is such that the time difference between the maximum storm surge and significant wave height varies with the typhoon condition. The mean time difference ranges from −2 to +1 hr, being related with the location in the bay. The mean storm surge and wave duration is respectively about 10 and 15 hr in the past typhoons and about 5 and 10 hr in the model typhoons. The time difference being 1 hr, the maximum wave-overtopping rate of a typical seawall may become less than half, the amount during a storm near a quarter. As mentioned above, the time difference and the duration are very important to evaluate the performance evaluation of coastal defense facilities.

Analytic expressions are derived concerning vertical circulation in tidal estuaries. Two mechanisms which may counteract are treated: density driven circulation and undertow in tidal waves. The paper is a “twin paper” to be read in combination with Bakker and Barber (2002). Focus of last-mentioned paper is: undertow. The present paper deals on density driven circulation. A first comparison is made with measurements and numerical computations.

The fight against oil pollution or the adequate forecasting require, among other things, a simulation “tool” to help taking the best decisions to preserve the harbour/coastal environment. This simulation toolkit must be able to forecast the position and drift of the oil spill and its physicochemical property evolution as a function of the spill features, the domain geometry and the atmospherical forcing. This paper presents the development of such an approach in the northern Spain where a suite of numerical models has been adapted and validated using field observations.

The objective of this paper is to describe the hydraulic climate of the Southern North- Sea by statistical models and physical laws for wave heights, wind setup, wave periods and wind speed. The focus is on the selection and calibration of parametric physical models for the use in the description of the joint probability distribution of hydraulic loads. For the validation and calibration of these models, observations have been used. An application is presented about the Rotterdam harbour extension.

Water quality and renovation inside harbour domains is becoming an increasing problem for many ports in the Spanish Mediterranean due to the augmentation of commercial and leisure activities inside the harbour domain. A second entrance is under construction in the Barcelona harbour, and the main aim of this paper is to assess the efficiency of two harbour entrances in enhancing the water renovation and quality inside the domain. Two numerical models developed at the Maritime Engineering Laboratory LEVI/UPC have been employed to evaluate changes in the circulation pattern.

The interaction of waves and currents at an inlet entrance can be significant. Traditionally, numerical modelers have separated the processes of tidal circulation and wave transformation, but the surf zone and inlet are areas where the interactions are strong and should be numerically simulated to capture the resulting hydrodynamics. This paper describes performance of coupled wave and circulation models for both an idealized inlet setting and an application for Grays Harbor, Washington, concentrating on the influence of waves on currents. A comparison of tidal current simulations to tidal-plus-wave-induced current simulations shows that the interactions create gyres, longshore currents, rip currents, and "shadow zones" of relatively weak currents. It is concluded that accurate simulation of the hydrodynamics at coastal inlets requires coupling of wave and circulation models.

The Ministry of Transport and Public Works in The Netherlands is investigating the required discharge capacity for sluices in the closure dike between the IJsselmeer and the Waddenzee to account for the effects of expected climate change. To that end, the water level on the seaward side of the discharge sluices has been predicted using a neural network for various scenarios for sea level rise and variations in the wind speed and -direction due to climate change. This will enable the effectiveness of various sluice designs and discharge strategies to be evaluated. The paper describes the design of the neural network, the preprocessing of the measured data and the training of the network. Training was carried out using only a part of the available data. Comparisons are then presented showing the quality of predictions for periods not included in the training. The results show a good agreement and justify the application of neural networks for this sort of application.

Long-wave oscillations or seiches observed in many harbours can cause damage to mooring and normal harbour operations. It is customary to explain and analyse seiches forced by free long waves outside the harbour or subharmonics generated by wave-wave interaction. Although the existence of a current along a coastline is not uncommon, typical seiche calculations overlook the influence of a current. Harmonic excitation due to an air jet is a widely read subject in acoustics. Following the analogy from acoustics, Fabrikant (1995) has proposed the possibility of harbour excitation due to a current. This concept is revisited in this article and the feasibility of seiches due to a coast-parallel current is discussed. The most important outcome of accounting for a long-shore coastal current outside the harbour entrance is that infinitesimal disturbances can grow exponentially to cause seiches even in the absence of external forcing.

This paper discusses the stability under wave action of breakwaters, seawalls or harbour walls formed by un-grouted stone or concrete blockwork. The loading on blockwork caused by the interaction between wave action and 'internal flows' in the hearting of the breakwater is calculated. Two loading cases were investigated for two different breakwaters: loading by wave impact; and loads at wave run-down. For the situations simulated, the response to wave impact loads was more or less independent of the permeabilities of the exterior blockwork and of the hearting or core of the structure. Loading during wave rundown is shown to be highly dependent on these structural parameters.

Entrained air in the form of small bubbles, or as larger air pockets, is thought to play a key role in determining the rise times and magnitudes of pressures exerted on coastal structures. In this paper simple equations for an aerated flow are presented. The equations are solved numerically and applied to the problem of pressure pulses in thin cracks or fissures in sea walls or breakwaters. The possibility of large pressure amplification in the cracks is demonstrated.

Large-scale model tests were carried out in the Large Wave Channel (GWK) with a slender cylindrical pile located at the end of a 1:10 slope. The test cylinder was subject to both regular and irregular waves. The paper deals with (i) the evaluation of the breaker characteristics at the breaking point for the test conditions, (ii) the comparison of the results with published data from small-scale experiments and (iii) the measured vertical distribution of horizontal particle velocities under a breaking wave. Additionally a new method of separating the measured force history into a slowly varying quasi-static and a dynamic part is presented by using the EMD (Empirical Mode Decomposition) for the analysis of breaking wave attack.

A numerical model has been developed to describe wave absorption, reflection and transmission arising from the interaction of long waves with a porous breakwater. The structure consists of a closely packed group of hollow tubes stacked with axes parallel to the bed and aligned in the direction of wave propagation. The model, based on the work of Sollit and Cross (1972) and developed for the limiting case of long waves, produces satisfactory agreement with experiments performed in a laboratory flume for a range of wave frequencies. Results from tests in the same facility on a horizontal, perforated plate device indicate that this form of wave absorber is extremely effective, particularly for relatively short waves.

Perforated wall caisson breakwaters will be used more frequently not only in relatively calm seas but also in outer seas in this century. The reflection performance is usually evaluated by hydraulic model experiments or some theoretical method. However, VOF numerical simulation developed by a specially organized research group in Japan can be used to evaluate precisely the reflection performance of the perforated wall caisson. The computer code named "CADMAS-SURF: Super Roller Flume for Computer Aided Design of Maritime Structures" was open for general uses. This report explains the procedure of the calculation and the comparison with the experimental and theoretical data.

Physical model tests were carried out on a 1:100 model of a typical Italian caisson breakwater, subject to a multidirectional wave attack causing mainly spilling and occasionally plunging breakers, from which a statistical description of breaker impacts is produced. The paper analyses the impact identification procedure and describes the statistics of the impact events exceeding quasi static wave load. The average impulse results independent from impact duration, whereas the variance increases as impact duration decreases. Most impacts last from 5 to 25% of peak wave period T_p, the average rise time being approximately 0.05 T_p. Local loading due to breakers does not appear to be conditioned by the wave field originating them, depending mostly on breaking wave conditions.

Larger turbines have made it economical attractive to produce wind energy for locations offshore, and during the last few years focus has been put on installing wind turbines offshore. The offshore locations introduce wave loads on the foundations which need to be accounted for in the design. In the present paper the wave induced loading on wind turbine foundations in Danish waters is investigated based on physical model tests and a Morison approach. For a typical foundation configuration the wave loading is inertia dominated, and the results show that good agreement between the measured forces and the Morison equation is achieved, when using C_D and C_m values which compare to the standard values.

The present paper shows results of the wave forces from the hydraulic model tests carried out in a big channel on a new design of concrete caisson to be used as vertical breakwater. This caisson has a rectangular inner "yard" much larger than the rest of cells located around it. This yard possesses, as main feature, the absence of foundation base to reduce the up-lift pressures, being able to be filled with a quarry stone or a dredging material. The usage of this dredging material could be considered an advantage for this new design of caisson from the environmental point of view. The caisson was subjected to four wave storms, each one with a different peak period, and several significant wave heights.

The paper describes the use of two fundamental design parameters, the void porosity and layer thickness in rock armour constructions. These design parameters are very sensible for factors such as the boundary definition of a rock layer, rock production properties, intrinsic properties and construction properties. Differences in the value of the design parameter cause a considerable financial risk. This risk contains two directions, the first is the effect on the hydraulic performance of the structure and the second is its relation to materials procurement. This paper describes and investigates the second risk for the contractor: the large margin in the calculation of the void porosity, which influences the amount of rock (in weight). Often this risk is on account of the contractor, so it is necessary to have better insight into these values in order to reduce the financial risk. Focus is on the variation of the porosity at the bottom, at the top and at the transition between two layers of graded material. This has resulted in correction coefficients for the layer thickness, as well as for the computation of the void porosity as a basis for payment by the client.

For many years, coastal engineers have used computational modeling tools based on continuum mechanics, to model fluid flow and to a lesser extent solid structure deflections and deformations. However, problems in coastal engineering are often concerned with the movement of discrete bodies and there is an entirely new breed of discontinuum modelling methods (DEM, FEM/DEM and DDA), designed to simulate body interactions. This paper considers the range of particulates that concern coastal engineers, discusses DEM applications generally before reporting on some modelling developments showing dynamically interacting realistic-shaped bodies in which the fluid behaviour is neglected. These numerical modelling developments are based on the combined discrete finite element method (FEM/DEM). The paper concludes with a brief discussion on future progress and opportunities.

It is recognised that thickness and void porosity of armour layers in rock armoured structures has a significant effect on the details of measurement and payment in contracts. Data from full-scale structures and model replications are examined to provide guidance on the range and average values of thickness and porosity that might be expected for a given Dn₅₀ of armourstone. The paper then takes a detailed look at the relations between the armourstone shape characteristics, construction methods and the resulting thicknesses and void porosities. The results presented are from a comprehensive set of dry model construction tests designed primarily to explore the effects of block shape rather than to model a specific breakwater or revetment. These generic tests introduce the "blockiness coefficient" to help quantify the most influential block shape effects. Results are compared with findings from full-scale structures and it is concluded that both laboratory and full-scale coefficients for void porosity and thickness can be adequately predicted provided the shape of rock armour pieces is quantified with the aid of the blockiness coefficient as well as the axial ratio (1/d).

This paper examines the hydraulic performance of armour layers made up of individually placed, tightly packed rocks. It uses the results of physical model tests to examine the ability of such layers to resist damage from wave attack and to limit overtopping discharge. Single and double layers are investigated. The performance of the structures is compared with the predictions of Van der Meer's stability and overtopping formulae.

Many coastal rock structures consist of multiple layers of graded high quality rock, placed on prepared foundations. Innovative structures with simpler cross sections of local rock placed on minimal foundations have however proved successful in a number of situations. This paper provides a review of existing design guidance, describes present practice and identifies advantages and constraints of innovative structures. The paper considers the predictability of, and selection of appropriate requirements for, performance with reference to other scheme elements. Finally, the conclusion presents opportunities for the implementation and further development of innovative coastal rock structures, including suggestions as to situations where such structures are most likely to be appropriate.

The present study features on the detail analysis of wave force and the stability of armor units of Rubble Mound Breakwater (RMB) through a numerical model. A two-dimensional numerical reshaping model of uniform sphere units' breakwater was developed to illustrate the phenomenon.

Calculation results show the time history of change of water surface, pressures, velocities, wave-induced forces, and the displacement of armor units. Significant relation between wave induced force distribution on the surface armor units and the surface elevations of wave at certain phase has been found. It is understood that the resultant forces during wave rundown are greater than during wave runup. The resultant forces are greater at the surface units located near the water surface. The model also shows the relation between the displacement area and water line oscillation on the slope. The effective area of dislodged armor units is found to be between the maximum runup and the maximum rundown. The maximum depth until where the effectice wave induced forces work onto the armor units is between 1.5 to 2 times of wave height. The influence of wave height on the wave induced force magnitude is not unique, but in combination with other factors, such as the slope angle. Under the same attacking wave characteristics, the steeper the slope of rubble mound breakwater, the higher the quantity of the dislodge armor units.

This paper describes development of a numerical simulation for predicting deformation of rubble mound seawalls. In the numerical method, VOF method is used for predicting wave field around rubble mound seawalls and DEM is used for predicting deformation. The calculations of VOF method and DEM are performed separately and alternately to take account of the interaction between fluid motion and deformation of the seawalls. Erosion-type deformation and accretion-type deformation are calculated with this numerical method. The numerical model makes a relatively good estimation of erosion-type deformation.

A synthesis of existing models and formulas is made to compute the virtual performance of rubble mound structures in shallow water under combined storm surge and breaking waves for sequences of hurricanes. The computed results for ten 500-yr simulations are presented for a typical structure as an example. The crest height and armor weight of the structure are designed against the peak of a 100-yr storm. The structure designed conventionally is exposed to approximately 350 storms for each 500-yr simulation. The computed wave overtopping rate and volume during the entire duration of each storm are analyzed to assess the severity of flooding hazards. The computed progression of damage to the armor layer is caused episodically by several major storms but slows down as the structure ages. The computed results are also used to quantify the equivalent duration of the peak of a storm which yields the same overtopping water volume and damage increment as those computed for the entire storm duration.

An abrasion model to estimate the loss of stone volume due to abrasion is applied to six existing Icelandic breakwaters, five of which are of berm type and one is a conventional breakwater. The characteristics of the stones have been determinated through extensive fieldwork presented here. An empirical correlation between abrasion resistance index (requested in the application of the model) and the point load index has been assessed. The results from the model shows that the volume reduction due to abrasion because of stone movements is only of the order of 1 % and therefore not significant.

The stability of rock slopes subject to breaking waves over a shallow foreshore is addressed in this paper. Physical model tests have been conducted on a 1:100 foreshore slope with a wide range of wave conditions including single and double peaked spectra and severely breaking wave conditions. The measured damage levels are compared to the Van der Meer design formulae, which were developed for conditions with no breaking on the foreshore. It has been found that the results of the design formulae are sensitive to the choice of input parameters (H_1/3 or H_m0 and T_m or T_m-1,0) and that the inclusion of the reduction factor for breaking waves (H_2%/H_s) in the design formulae should be applied only with caution, depending on the choice of input parameters. These results apply for the conditions tested and should be verified for other foreshore configurations before a definite recommendation pertaining to the use of the Van der Meer formulae for shallow foreshores can be given.

Studies on the stability of the amour layer (d'Angremond et. al. [1999] revealed the importance of density of placement. The current research focuses on the influence of the density of placement on the stability of cubes in a double armour layer and Tetrapods and rocks in a single armour layer. The experiments were performed in the Laboratory of Fluid Mechanics of the Faculty of Civil Engineering and Geosciences at the Delft University of Technology. A model of a breakwater was constructed in a wave flume. The density of placement of various types of armour layers was varied. An increase in density of placement resulted in all cases in an increase of stability, except in the case of cubes. Cubes have the tendency to start behaving like a placed block revetment including the characteristic failure mechanisms like uplift and sliding. Tetrapods seem to be unsuitable for single layer armour layers due to the fact that the filter layer is easily attacked by the waves even when no tetrapod has been removed. Experiments on rock as armour material showed that vertically placed elements lead to a much more stable construction due to their self-repairing ability. Characteristic for rock is the piling up of elements under the waterline caused by the impact of collapsing waves. This leads to very low densities of placement higher on the slope. Existing damage criteria are not suitable for density of placement. In this research an effort has been made to create damage criteria, which consider area of attack, density of placement and different failure mechanisms.

Recent research has shown that it is possible to use concrete cubes as a single armour layer on breakwaters. However, this is only possible in case a placing density of at least 70% is achieved. In lab-tests it is no problem to place cubes with this density, however it is questionable if such a placing density can be achieved in prototype situations. A number of tests have been carried out in order to determine the requirements for dropping concrete blocks from a crane onto a breakwater slope. The conclusion is that at a waterdepth of approximately 10 times the block size this just can be achieved and that at smaller depth this placing density can be reached without too many difficulties. However, in all cases a good quality toe is vital for a high placing density. Blocks have to be placed at an angle of 45° with respect to the breakwater axis; blocks placed parallel to the breakwater axis will lead to large voids.

A monitoring program of dolosse on the Crescent City, California, outer breakwater has produced 15 years' worth of data, including static stress and concrete strength measurements, movement and translation data, breakage surveys, and wave climate statistics. The latest field study, conducted in 1999, had three primary goals: (1) instrument verification and in situ static structural response; (2) concrete strength testing of dolosse near the cap and still water level; and (3) an inventory of the extent and type of dolosse breakage. The results of the 1999 field study will be presented in this paper and long-term trends will be identified and evaluated.

This paper discusses recent experience with the application of the Core-Loc^® armor unit. Practical issues associated with implementation of Core-Loc^® as the armor unit on breakwaters are discussed, including stability, placement technique, packing density, the toe and the crest details.

A new deepwater port development at Punta Langosteira near La Coruña is under design. Shelter is provided by a 2 km long rubble mound breakwater fully exposed to Atlantic waves. Max. water depth is 45 m, crownwall crest level is + 25 m and main armour blocks are 150 t cubes. Moorings for oil tankers are arranged along the inner side of the breakwater. The paper presents the design criteria, the design procedure, the main results from model testing, and the subsequent reliability evaluation and optimisation of the cross section design for the most exposed part of the breakwater. Model test results related to armour stability, overtopping, and wave wall forces are compared to predictions by available formulae.

The aim of this paper is to define and generate the method of predicting a characteristic pore velocity in the core of rubble mound breakwaters under wave attack. This characteristic pore velocity is required to properly scale the material used in scale model tests in the lab. The authors propose the RMS value of the velocity averaged in four control points as "characteristic pore velocity" and provide a comprehensive method to predict this velocity in prototypes.

The main objective of this paper is to study the attenuation of the wave induced pore pressures inside the core of a rubble mound breakwater. The knowledge of the distribution and the attenuation of the pore pressures is important for the design of a stable and safe breakwater. The pore pressure attenuation is studied using experimental data (in Section 2) and using a numerical wave flume (in Section 3). The experimental study includes the analysis of large scale data from a physical breakwater model and prototype data from the Zeebrugge breakwater. The large scale data are taken from literature and have been re-analysed in detail with respect to the attenuation characteristics. The analysis follows the method by Burcharth et al. (1999) and confirms the practical calculation method for the attenuation of the pore pressure in the core given in this reference. The attenuation of pore pressures in a breakwater core is also studied in the numerical wave flume VOFbreak². The numerical results are compared to the results from the experimental study.

An advanced design method for armor units covering a rubble mound foundation of a composite breakwater is proposed. The method is based on a wave induced flow field near the breakwater mound, which is calculated by a numerical wave flume CADMAS-SURF. Trial computations and comparisons with experimental results demonstrate the validity and usefulness of the proposed method.

The breakwater at Nafanua Harbour, 'Eua, Kingdom of Tonga was overtopped and damaged by an raised sea elevation "surge", combined with large swells from 300km distant Tropical Cyclone Paula. Initially the cause of the surge was not clear because the Cyclone was distant. Measurement of runup debris lines combined with estimates of swell size and wave setup on the fringing coral reef platforms indicate that an unusually large wave setup was the main reason for the overtopping event.

This paper discusses a number of applications of digital image technology in port operations and coastal monitoring. Some of these applications were originally applied in physical scale models of harbours and coastal structures, at the CSIR laboratories in Stellenbosch, South Africa. More recently, these were applied to different monitoring exercises along the South African coast, and specifically in and around the major ports.

This paper is devoted to the numerical simulation of wave propagation and wave diffraction by an immersed structure in viscous flow. An original approach using a diffracted flow defined as the difference between total and incident flows is followed. The incident flow is defined explicitly using nonlinear potential flow theory; Navier-Stokes equations and nonlinear free surface boundary conditions are solved for the diffracted flow only. This procedure, which is very efficient in terms of computing time and accuracy, was first developed by Ferrant (1996) for 3D nonlinear wave-body interaction in potential theory.

Wave fields around a breakwater head with wave overtopping were investigated by using large eddy simulation. The complex forms of vortices and Lagrangian vortical fluid motion involved in the vortices were discussed. The three-dimensional eddy structure around a breakwater head is composed of many vortex tubes intertwined and interacting with each other. A complex eddy structure around the breakwater head might cause local scour and promotion of sediment suspension.

The relative importance of the different velocity moments in sediment transport around detached breakwaters was investigated by using Bailard's (1981) formulae. Current velocities were measured during a laboratory experiment on shoreline evolution behind detached breakwaters, conducted in the UK Coastal Research Facility at HR Wallingford. Regular and irregular unidirectional waves were simulated. Normalised moments were calculated and compared to predictions from simple monochromatic and linear random wave models. A reasonable agreement between observed values for irregular waves and the predictions of a random linear wave model was found. Third and fourth order velocity moment terms from Bailard's formulae were decomposed into a number of velocity moments which predict bedload and suspended load rates related to the contribution of mean flow, short and long waves and their interaction. The contribution from short wave stirring and transport by mean flow was found to be dominant for regular and irregular waves. In the case of irregular waves this contribution was almost matched by contributions from stirring and transport by short waves and from stirring by short waves and transport by long waves. It is clear that the accurate prediction of velocity moment terms and related sediment transport requires the consideration of the contribution of mean flow, short and long waves and their interaction, and hence irregular waves.

This paper reviews existing methods to evaluate wave loading on exposed jetties and related structures. This work aims to address inconsistencies and gaps in these methods. It was motivated by the increasing demand from UK designers and contractors for clear guidance on prediction formulae for the hydraulic design of these structures, and particularly for guidance on wave forces on decks and beams. A series of flume tests to measure wave-induced forces on a model of a jetty structure was carried out at HR Wallingford. Horizontal and vertical wave forces on beam and deck elements were measured. Initial results and comparisons with some of the existing models for evaluation of wave loading on jetty structures are presented here. The paper provides initial guidance on wave forces on exposed jetties and fills gaps in present design methods through analysis of these new experimental data.

While the conventional decomposition of incident and reflected waves in the frequency domain analysis employes the Fourier transform, the time domain decomposition proposed here ultilizes the Hilbert transform. One of the advantages of the time domain decomposition is that the dispersion relation of linear water wave is no longer pre-required. It then makes feasible the decomposition of long waves in nearshore region where free long waves and bounded long waves by wave groups are usually co-existing.

In this paper, we examine several methods for calculating the reflection of irregular waves from a perforated-wall caisson breakwater using a regular wave model. The first method is the traditional regular wave approximation, which approximates the irregular waves as a single regular wave whose height and period are the root-mean-squared wave height and significant wave period, respectively. The second is to use the regular wave model, repeatedly, for each frequency component of the irregular wave spectrum. The wave period is determined as the reciprocal of the frequency, and the root-mean-squared wave height is used for all the frequencies. The third method is the same as the second one, but the wave height corresponding to the energy of each component wave is used. Comparison with experimental data of previous authors shows the second method is the most adequate, giving reasonable agreement in both overall reflection coefficients and reflected wave spectra. The first method, which uses the most conventional approximation of irregular waves and has been widely used because of its simplicity, gives large errors even in the overall reflection coefficients, leaving aside the fact that it cannot resolve the frequency-dependent nature of wave reflection.

A preliminary investigation of scour around a horizontal cylinder is presented in this paper. A coupled computational fluid dynamics and sediment transport model, FLOW-3D, was used to model the velocity around five static bed profiles, and the dynamic scour around a cylinder. Laboratory data from Jensen et al. (1990) was used to evaluate the modelled bed velocity. Comparisons are reasonable, although the bed velocity in the region of the cylinder is under-predicted. Modelled scour evolution were then compared with laboratory data from Mao (1986). The model predicted the general size and shape of the scour hole however there was an over-prediction in the length of the equilibrium scour hole. The dune created downstream of the cylinder is more prevalent than is present in the laboratory data. The model also overscoured upstream of the cylinder. Time scale to reach equilibrium was found to be on the order of 50 times shorter than the laboratory results.

Installing submerged dikes on both sides of channel and basin has been confirmed as an efficient method for siltation protection. This study is intended to find most efficient cross-section for siltation protection, and several kinds of cross-section of dike have been tested through laboratory experiments and numerical simulations. Tests have been conducted using Slim-tank, which is a special flume to enable visual observation of fluid mud movement under several conditons such as wave existing or current existing, and finally most effective cross-secition of submerged dike has been proposed.

Near-bed structures are studied in this paper, in particular submerged rubble mound structures with a relatively low crest such that wave breaking does not have a significant influence. Physical model tests have been performed to study the stability of near-bed structures under wave loading, with or without a current. Based on the new data and an analysis of existing data, several methods to predict the stability of near-bed structures have been analysed. One of these methods was found to be the most appropriate. This method was calibrated to relate the erosion of near-bed structures to a mobility parameter. It was found that for low-to-moderate currents in combination with waves, the waves dominate the stability of the rock material; the stability of the near-bed structures can be predicted without taking the influence of the (following) current into account. The obtained prediction method is simple and accounts for the effects of wave height, wave period, number of waves, stone diameter, rock density and crest elevation. More detailed information can be found in Van Gent and Wallast (2001).

This paper presents a numerical wave tank (NWT) approach of studying wave-structure interaction. A general energy conservation equation is derived to describe total energy variation within the NWT. The equation is then used to define wave transmission, reflection, and dissipation coefficients (e.g., K_t, K_r, and K_d). The realistic wave interaction with various offshore and coastal structures is simplified in this study as the solitary wave interaction with an impermeable rectangular body (a × b). Based on a series of systematical numerical experiments, some empirical formulas are found, which express K_t, K_r, and K_d as the individual functions of relative wave height (H/h) and structural height (b/h). The current study serves as the first step to establish the ultimate formulas for predicting wave coefficients as the general and composite function of H/h, b/h, and a/h, e.g.,K_t,r,d = K_t,r,d(H/h, a/h, b/h).

Shore parallel low-crested structures are potentially interesting but very difficult to design "functionally". The functionality of these structures is sensitive, in particular, to mean sea level variations, which means that freeboard varies and, therefore, that the structure will behave as expected only under a limited range of conditions. The same applies to the incident wave transmission (and even reflection) since the filtering effect is a function of wave height and period (length).

Increasing demands for soft shore protection often lead to the application of artificial reefs as an active measure already decreasing the waves before they reach the shore. To make a better use of the advantages of this type of structure the physical processes at the reef have to be understood. With this background physical model tests on artificial reefs have been performed at Leichtweiβ-Institute for Hydraulic Engineering utilizing flow visualisation to get an insight in the processes contributing to the hydraulic performance of artificial reefs with rectangular shape.

Time series of regular and irregular water surface elevations from various experimental configurations are examined with time series analysis methods and Fourier analysis techniques in a laboratory wave flume concerning the hydrodynamic wave damping performance of an artificial reef. The decomposition phenomenon of nonlinear wave trains passing over an artificial reef has been investigated experimentally. It is demonstrated for monocomponent wave trains that the spectral energy distributions and the total amount of energy within the spectra is significantly dependent on the position of a wave gage in the wake of the reef since all Fourier based analysis techniques are incapable to identify nonstationary processes. The Hilbert-Huang Transformation (HHT) is chosen to operate as a more appropriate tool to demonstrate the nonlinear processes of the transmitted transient waves induced by the presence of the artificial reef. Results of this technique are shown and prove that the nonstationary sea states behind the structure are essentially composed of individual dispersive waves which are phase shifted in the wake of the struture and propagating individually to generate a beat effect. Further experiments with irregular waves (JONSWAP) are also carried out and analysed with the HHT.

The present paper presents an experimental and numerical study of partial wave reflection and transmission for monolithic rectangular submerged breakwaters located in the nearshore region. The study is based on large-scale experiments that have been conducted in Universitat Politecnica de Catalunya, Laboratory de Ingeneria Maritima (LIM-UPC). The most important conclusion, deduced by the experiments, is that the main energy dissipation mechanism is the wave breaking due to the presence of the structure. The numerical approach of the problem is based on a higher order Boussinesq type of equations. The comparison between model results and experimental data shows that the model is able to describe breaking and non breaking wave deformation passing over a submerged breakwater.

This paper describes a numerical model for the simulation of near shore wave dynamics and bottom topography change. In this paper, the nearshore wave dynamics is simulated by solving the depth integrated Boussinesq approximation equations for nearshore wave transformation together with the equation of continuity with a Crank-Nicholson scheme. The wave runup on beach is simulated by a scheme similar to the Volume Of Fluid (VOF) technique. The wave energy loss due to wave breaking and shear generated turbulence is simulated by a k-ε model, in which the turbulence kinetic energy (TKE) generation is assumed as the sum of those respectively due to wave breaking and horizontal and vertical shear. The bottom topography change model calculates the transport of bed load, the uptake, advection, diffusion and deposition of bed materials under combined actions of waves and currents. Especially, a new numerical scheme based on the cell donator-receiver concept has been introduced for the discretization of terms representing the advective transport of suspended sediment. Also, a new suspended sediment pick-up function, which relates the sediment pick-up rate with bed shear stress and turbulence intensity has been introduced. Additionally, a new numerical scheme, which is also based on the cell donator-receiver concept, is proposed for the evaluation of the contribution of bed load transport to the bottom topography change. This new numerical scheme eliminates the necessity of introducing false diffusion terms to the equation governing the bottom topography change, as done by various other researchers, including Rakha et al (1997), Watanabe et al (1988), and thus enable an accurate prediction of the bottom topography change, especially near coastal structures. The verification of the numerical model against data obtained from various indoor experiments reveals that the model is capable of simulating the wave dynamics, turbulence and bottom topography change under wave actions. Especially, with the new differencing scheme for the bottom topography change equation, the numerical model is capable of simulating the wave induced scouring near coastal structures with acceptable accuracy and relatively short computational time.

The purpose of this research is to provide a detailed investigation on the scour near coastal structures by collecting and analyzing indoor experiment and field data and by a numerical model; and based on it to establish a quantitative method for the evaluation of the maximum scouring depth.

Applicability of remote sensed images to monitoring long and short term beach deformations has been investigated. Long term shoreline displacements analyzed by satellite imageries were comparing with field survey data in the corresponding period. The agreements are found to be satisfactory, thus the present analysis using satellite imageries are validated. Next, short term beach evolutions induced by typhoon attack are discussed. Most of the obtained characteristics, which are considered to represent typhoon induced shoreline changes, are found to be physically reasonable. Also, the succeeding shoreline change suggests that beach recovery process occurs after the typhoon-induced erosion. Furthermore, to explain the physical process of the typhoon induced shoreline change, numerical analyses on the wave hindcasting and wave transformation have been conducted. The results of typhoon hindcasting and shallow water transformation reveal that even a small change in wave incident angle results in a great difference in the wave height distributions in the alongshore direction of the study site.

A new coastal development was proposed on the coastal reef flat of the Red Sea, just north of Jeddah. The site chosen for this aquarium development occupies a part of the reef flat not previously reclaimed. The reef, just 25m seaward of the site, drops extremely steeply (typically 4 vertical to 1 horizontal) to over 200m depth. The absence of shallow water up to the reef allows deep water waves to approach unrefracted and unbroken from the main westward sectors. The paper presents the design problems and dual approach methodology of using physical model results to validate an empirical model of wave loadings and reactions to optimise the design. It provides details of how the traditional design methods were modified and extended to incorporate steep bed slopes and protective breakwaters. This includes comparisons of measured and predicted methods with respect to armour and caisson stability.

Within Dun Laoghaire Harbour on the East Coast of Ireland a new 650 berth Marina was constructed during 2000-2001. The scheme required the construction of two breakwaters each 450m long with 3.5-7t protection rock armour on the outer face and with the main eastern breakwater having a 5.5m wide promenade. The original design for this was a traditional caisson crown wall breakwater. This paper describes the development and construction of an alternative design using reinforced earth principles to form a more cost effective crown wall system.

A numerical simulation model on coupled motions of waves and a submerged floating structure is proposed in the present paper. So far, this type of numerical models was based on a potential theory under assumptions of inviscid fluid and irrational fluid motion. The governing equations of the present model consist of the Navier-Stokes equation and continuity equation. The Arbitrary Lagrangian and Eulerian method and the boundary fitting coordinate method are applied. Free surface displacement near the floating structure, wave pressure acting on it and displacement of the structure are compared between calculation and physical model tests to verify the present numerical model.

New rapidly-installed breakwater technology for nearshore logistics is discussed. The breakwater concept consists of one or two fabric tubes filled with water or sand and pressurized. A fabric shroud is stretched over the top of the tubes and anchored to the bottom on both the seaward and leeward sides. A physical model study of the conceptual design showed that the technology is feasible. In the flume study, forces were measured in the fabric shroud. Incident and transmitted waves and sediment scour were also measured. It was found that the maximum wave-induced forces were less than the capacity of presently available screw-type anchors. Also, the wave transmission coefficients were slightly above those of rubble mound structures and within the anticipated range.

Functional design of low crested breakwaters requires an accurate prediction of wave transmission in the protected areas. Nevertheless, commonly used formulae do not appear to be reliable enough, especially for structures located in shallow waters. The paper describes results from large-scale model tests conducted on rubble mound breakwaters exposed to breaking waves. Tests were carried out at the "Grosser WellenKanal" of Hannover, Germany. The model dimension, near to prototype, allowed minimizing scale effects connected to wave breaking. Existing formulas on wave transmission have been verified and influence of crest width and breaker index have been highlighted.

A double-submerged breakwater, which is a pair of small submerged breakwaters placed some distance apart, is considered to obtain more desirable wave field than a conventional single-submerged breakwater. Wave tank experiments using random waves of several significant waves were conducted for double- and single-submerged breakwaters. Spatial changes of wave amplitude spectra and flow velocity spectra, mean water level change, skewness and atiltness of the wave profile of the water surface and the bottom velocity oscillations were calculated. The single-submerged breakwater generally shows better wave interception, while the double-submerged breakwater has less mean water level change, less steady flow velocity and less asymmetry wave profile of the flow velocity at the bottom. It is also found that very low frequency oscillation arises in surface waves and flow velocity at the bottom, especially for the single-submerged breakwaters.

A large reef has been constructed offshore from Narrowneck on the Gold Coast, Queensland, Australia. The reef provides a submerged, low visual impact, coastal control point to stabilize the nourished northern Gold Coast beaches. As the reef is in a popular tourist and surfing area, it has also been designed and constructed to enhance recreational amenity. For safety and cost efficiency, the reef has been constructed of very large sand filled geotextile units. Experience with the design and construction of such structures is limited and the inclusion of improved surfing as secondary design criteria increases the complexity. To facilitate evaluation and modifications to the reef there is a comprehensive monitoring program, which includes the use of ARGUS video imagery to assess the changes in the shoreline related to wave conditions at the wave rider buoy nearby. Divers are studying the physical performance of the sand filled geotextile containers, as well as the extent and diversity of the marine ecosystem which has exceeded expectations. There have been a number of storm wave events during the monitoring period. The data obtained from the monitoring is being used to modify the long term Narrowneck reef shape, and to design other proposed reefs in the area.

This paper presents the construction, surveying and results of a prototype experiment on stability of rubble mound protection for submarine outfalls. Taking advantage of the Santander outfall construction, financial support from the European Community was obtained to carry out the experiment that consisted of covering the outer layer of the outfall's protection with two extra layers of rubble with a stone weight smaller than that of the project. Three stone weights were tested in three different stretches of the outfall. The experiment lasted two years, during which waves and stone movements were surveyed. Damage and wave data were analyzed and the damage parameter is presented here as a function of the mobility parameter. To take into account the random characteristics of waves in the sea states, a Montecarlo simulation is used here to calculate the mobility parameters corresponding to all the waves of all the sea states that reached the experimental sections. The use of the average of the 50 biggest mobility parameters that reached the test sections at the survey date is proposed here to represent damage results against the mobility parameter. Using this mobility parameter, prototype results compare well with laboratory experiments carried out with regular waves.

In 1995 The Department of the Environment (DoE) commissioned HR Wallingford (HR) to learn from experience the essential elements of barrage design and consolidate the knowledge into guidelines for future application. The target audience is local authorities, the Environment Agency, development corporations and others who may wish to promote a barrage scheme, and consulting and contracting engineers who will be responsible for the design and construction.

The objective of the Environment Agency contract was "to review available experience on the design, operation and environmental impact of estuarine barrages in order to provide best practice design and operational engineering solutions to overcome or mitigate problems and enhance, where possible, the aquatic and riparian environment".

The paper describes the guidelines, highlighting the key issues form both engineering and environmental perspectives. The guidelines recommend that future barrage schemes should closely embrace the principles of sustainable development, thus ensuring that economic investment and environmental improvement go hand in hand.

The issues covered by the guidelines are: fisheries and conservation, water quality, hydrodynamics, morphology, flood defence, groundwater, navigation, waves, and structural design. Throughout the research, gaps in knowledge and technology were identified and a list of research priorities is given.

The Cardiff Bay Barrage impounds the estuaries of the rivers Taff and Ely in South Wales. It creates a 200 hectare lake and 12 km of attractive waterfront. The 1100-m-long barrage includes sluices, locks, an outer harbour, fish pass, embankment and bridges. The sluices have a capacity of 2300 m³/s, to discharge river flood flows. The gates close to exclude the highest tides, allowing bay water levels to be controlled even in extreme conditions. The Cardiff Bay Barrage has many innovative features that combine to create a unique project. Among these are:

• construction in the second highest tidal range in the world

• largest European fish pass

• new filling method for locks in hypertidal estuaries

• float-in caisson breakwaters with double-wall "Jarlan" cross-section

• unique integration of engineering and architecture

Cardiff Bay was impounded with seawater in 1999, and has achieved its objectives. Sluices have been subject to high river flows, the maximum being 535 cumecs when the Bay level rose by 2.2 m as forecast. The locks have accommodated up to 230 boats passing through in one day. Fish numbers for 2001/2002 are in line with expectations. The saline sump is designed to contain salt water that enters the Bay via the locks, whilst the flushing pipe discharges sea water at low tide. At times the containment of sea water proves problematic. Alternative methods to avoid lock restrictions are being examined.

During the course of the conception, approval, design and construction of the Cardiff Bay Barrage many hydraulic studies were required at various levels of complexity and detail. HR Wallingford carried out many of these studies. The studies began with a pre-investigation feasibility study in 1985 in partnership with Wallace Evans and Partners, leading to a full feasibility study in 1987. When the contract for the design of the barrage was awarded to Sir Alexander Gibb and Partners further studies were required. Additional advice on the hydraulic aspects was also provided during construction from 1992 onwards. The studies involved field measurements and the integrated use of a number of computational and physical hydraulic models. This paper focuses on the sediment transport and wave aspects of those studies, giving a brief account of the issues, the studies and the main conclusions.

A Severn Tidal Barrage could generate 8.6GW of renewable electrical energy and save up to 18Mt of CO₂ per year. In addition, it would protect 140 miles of coast from flooding, it could provide road and rail crossings of the estuary, and it would stimulate investment in the region bringing employment and leisure benefits. Major studies of the project were carried out in the 1970's and 1980's and this paper concludes that a re-appraisal of the project should be carried out now.

This paper describes briefly the construction phase of a large headland system in a dominant unidirectional longshore sand transport environment. The project consists of a revetment, six large headlands and beachfill to protect a naturally eroding beach, to minimize the downdrift impacts through bypassing the historic supply of sand and compensating for part of the sand budget deficit, and to allow for continued seine net fishing. An intensive monitoring program has been conducted to test several critical and underlying assumptions of the design phase and to modify the design accordingly for this Design-Build project. Monitoring results are presented and discussed.

Full scale wave run-up measurements have been performed on the Zeebrugge rubble mound breakwater. Wave run-up also has been investigated on various small scale models of the Zeebrugge breakwater. A significant difference between the results has been noticed. Additional small scale model testing has been carried out on a slightly modified scale model: a regular armour unit pattern has been applied in stead of an irregular pattern as in full scale. The aim of the additional laboratory tests was to investigate the influence of the spectral width parameter ε and the influence of the position of the wave run-up step gauge with respect to the armour unit pattern and the water level.

Hydraulic characteristics of a new type of low crest seawall are investigated by laboratory experiments in terms of overtopping rate, distributions of wave height and mean water level, wave crest and trough levels at the vertical wall. The seawall consists of a wide submerged reef and a permeable terrace in front of a solid wall. The permeable terrace in the model is a wire netting box filled with gravels and the width is changed for 5 cases. The measured overtopping rate is compared with the overtopping rate calculated for conventional types of seawall. Results confirm that the permeable terrace with an appropriate width is very effective to reduce wave overtopping and produce a low crest seawall.

The wave transformation at vertical and sloping perforated walls (of slope cot α = 0 - 6 and porosity ∊ = 0% – 40%) has been investigated experimentally. A basicly linear concept has been developed for the analysis of the hydraulic processes. A set of predictive equations has been derived for wave transmission, wave breaking, wave reflection and wave energy dissipation on permeable slopes.

Cape Lookout State Park on the northern Oregon coast has experienced extensive erosion in recent years, much of it associated with the strong El Niños of 1982-83 and 1997-98, culminating with the series of unusually severe storms during the La Niña winter of 1998-99. The erosion removed the line of high dunes that had existed between the ocean beach and park facilities, so winter storms were able to flood the park grounds. While a shore protection structure was clearly needed, it was determined that a conventional quarry-stone revetment or seawall would not be used. Instead, it was decided to construct a dynamic revetment or cobble berm, one that would be similar to a natural cobble beach, backed by an artificial dune having a core of sand-filled geotextile bags, covered by loose sand and dune vegetation. Our study was initiated to monitor the effectiveness of these structures, including a program of periodic surveys, analyses of tides and wave runup compared with structure elevations, and the progressive changes in the structures wherein they evolve to become more like their natural counterparts on the coast. To date the structures have survived three winters, including major storms that produced some overtopping. The ultimate goal of the study is to provide improved design criteria for the use of such "natural" structures where conventional approaches to shore protection cannot be used.

The stability of revetments, consisting of regularly placed blocks and grouted with asphalt, is jeopardized by static uplift pressure due to tidal water level variations and by wave attack. The stability has been analyzed theoretically, with field tests and with large scale model tests in the Delta Flume of Delft Hydraulics.

In the Netherlands many placed block revetments have been strengthened by pouring asphalt grouting in the cracks between the blocks. This ensures that the revetment is less susceptible to wave attack. However, because the top layer has become relatively impermeable, the construction may well be susceptible to upward water pressures that build up in the filter layer beneath the top layer. To gain more insight into the physical processes and to be able to improve on the current conservative stability rules an infiltration test was performed on an existing asphalt grouted placed block revetment. This paper deals with the preparations for this test, the observed phenomena during the test and its significance for the engineering practice.

Recent reconstructions of placed block revetments extended the range of cover layer thickness up to 0.5 m. The equations describing the permeability of block revetments were only verified in the range from 0.1 to 0.25 m. So to extent the range to cover layers up to 0.5 m field and laboratory tests were performed and analyzed.

Low-crested breakwaters or dams may provide some sheltering of flood defence structures like dikes or seawalls. The area around such dams and flood defences very often has a complex geometry. This paper describes the process of transformation of extreme wave boundary conditions in front of such structures to the flood defence structures. Methods to calculate the required crest height for various types of complicated sea walls are given, based on allowable overtopping.

The interaction of wave overtopping and the soil properties of a seadike is responsible for the initiation of dike failures and the breaching process. Unfortunately, the present design of seadikes is based on the separate determination of hydraulic and geotechnical design parameters. Therefore, large scale hydraulic model tests were set-up on the basis of a detailed failure analysis for seadikes to investigate the infiltration and erosion process due to wave overtopping. Results are presented in this paper.

Wave overtopping was responsible for many dike failures in the past. It is not feasible to avoid wave overtopping completely in the future due to uncertainties in the prediction of the design water levels and the costs of uneconomical high dikes. Therefore, wave overtopping has to be taken into account for the design of seadikes. The overtopping flow velocities and related layer thicknesses are required which are responsible for infiltration and erosion of the dike crest and the landward side of a dike. The objective of this paper is to study the flow velocities and related layer thicknesses associated with wave run-up and wave overtopping based on theoretical and experimental investigations.

The Japanese western coastal areas were heavily damaged by storm surges and waves due to the Typhoon No. 18 on September 23~24 in 1999. The waves induced by the typhoon approach also were huge on the time when the storm surge became the maximum. Some coastal zones located on the west side of a target bay were perfectly flooded by the overtopped waves at the sea walls. The flooded area, inundation time and wave overtopping rate in the two districts were revealed by the field survey. One was protected by the sea wall covered with armor blocks and another by sea wall with a detached breakwater. A numerical scheme expecting wave overtopping rate was applied for investigating the effect of coastal barriers. The numerical results agreed with the flooded volumes obtained in the field survey and reproduced the reduction of wave overtopping rate in the detached breakwater.

safe use of low lying and densely populated coastal regions depends critically on the performance of coastal structures in defending these areas against storm surges, wave attack, flooding and erosion. Continuing sea level rise and climate change (storms are becoming rougher) emphasise the need for reliable and robust predictions as higher storm surges and bigger storms may lead to flooding. Population pressures on land use in coastal regions have sometimes ignored age-old appreciation of coastal hazards. The CLASH research project EVK3-CT-2001-00058 is being funded by the EU to provide "Crest Level Assessment of coastal Structures by full scale monitoring, neural network prediction and Hazard analysis on permissible wave overtopping". It is intended to produce generally applicable prediction methods based on permissible wave overtopping and hazard analysis. This paper describes the general approach of this major European project and more specific the development of a homogeneous overtopping database, which will be the basis for the general prediction methods.

Measurements at HR Wallingford have indicated that bimodal wave conditions may be a worst-case condition in the design of some coastal structures (Hawkes et al. 1997). This paper describes a simple method for predicting the mean rate at which embankments are overtopped in bimodal seas. Use is made of the H&R model (Hedges and Reis 1998). This model is more soundly based on physics than earlier procedures for predicting overtopping discharges. It has also been shown to provide the most promise for future development (Hawkes 1999). Its use in predicting mean overtopping discharges in bimodal seas is the latest development. Comparisons between measured rates and model predictions show generally good agreement over the wide range of conditions covered by the tests reported here.

Laboratory and prototype overtopping observations of the Zeebrugge breakwater were analyzed using artificial neural networks (NN). An evolutionary strategy was used to calculate a chain of two pruned NN models able to classify "significant" overtopping events (q > 10^-4.51/m/s) and to estimate overtopping discharges. This NN methodology proved to be effective for analyzing a small number of data (only 113 cases). Using the complete NN model as a simulator, thousands of virtual tests were generated from which conventional empirical formulae were obtained. Using the classical dimensionless overtopping variable Q = q/[g H_m0³]^0.5, an exponential formula agreed with the results using the following dependent variables: Rc/H_m0, Ir, Rc/Dn and U. More than 80% of the overtopping estimates fell in the half-twice interval of the observed overtopping rates, not as good as the NN model, but very good for most practical applications. The comparison with prototype overtopping measurements was inconclusive because of the lack of proper scaling laws for windspeed in the laboratory and the lack of moderate and high overtopping measurements in the prototype.

The majority of prediction methods for overtopping of seawalls are based on physical model tests under simple 2-dimensional conditions. There is some evidence (not unambiguous) that overtopping may increase at small degrees of obliquity, and that corners (in plan) may give local concentrations of overtopping. This paper, produced as part of the VOWS project on impulsive (violent) overtopping of vertical seawalls, describes experiments to measure mean and wave-by-wave overtopping discharge under conditions of oblique wave attack and at 3-d corners. Results analysed in the first phase of this 3-d study suggest that mean overtopping discharges reduce significantly with increasing angle of wave attack and that the occurrence of impulsive overtopping diminishes rapidly with obliquity of wave attack > 30°. It is also observed that overtopping may not increase in corners with an approach beach or berm.

Motivated by questions raised by developers of wave energy devices based on wave overtopping concepts, model tests have been performed to study overtopping of structures with limited draught, low crest freeboards and slope geometries designed to increase overtopping and thereby also the captured amount of wave energy. The test results have been incorporated in an existing overtopping prediction formula by application of correction factors presented in the paper.

Wave overtopping over dikes may cause dangerous situations, one of which is breaching of dikes. Wave overtoppmg can be characterised by mean overtopping discharges but this does not provide information on individual wave overtopping events. This paper focuses on parameters related to individual wave overtopping events with a low probability of exceedance. Measurements of velocities and the thickness of water layers have been performed at the crest and inner slope of dikes. The height, width and roughness of the crest have been varied and also the angle and roughness of the inner slope. Based on these tests, prediction formulae have been derived for velocities, the thickness of water layers, and low-exceedance wave overtopping discharges. The estimates of these quantities can be used to characterise the wave loading on the crest and inner slope of dikes, especially as the characteristic wave loading for the initial phase of breaching of dikes initiated by erosion at the crest or inner slope. Besides above mentioned quantities the measurements were used also to obtain more information on for instance mean wave overtopping discharges and on volumes of water within individual overtopping waves.

This paper describes new research under the VOWS (Violent Overtopping of Waves at Seawalls) project. Established guidance on admissible overtopping volumes is based upon values of mean discharge. In cases where hazard to pedestrians/vehicles are concerned, it is clear that an admissible level of overtopping would be more appropriately based upon the volume of an individual overtopping event. Further, the hazard presented is not only a function of the volume of that overtopping wave, but also of the speed and trajectory of the jet. Methods exist to predict maximum individual overtopping volumes. This paper presents new data and a first predictive tool for overtopping "throw velocities", and first data from a device designed to measure directly the trajectories of overtopping jets.

It has often been suggested that wave overtopping processes measured in physical models may include scale effects which could influence the reliability of any predictions to full scale. This paper describes studies at small- and large-scale under the VOWS (Violent Overtopping of Waves at Seawalls) collaborative research project, and presents detailed results on overtopping performance of 10:1 battered seawalls, at small- and large-scale. These results can be used to predict mean overtopping discharges and wave-by-wave overtopping volumes. Overtopping measurements are compared with predictions for vertical walls. Results quantify mean overtopping rates and characterise the uncertainties of scale effects in overtopping measurements. Within experimental limitations, the results show no significant difference in small and large scale overtopping, demonstrating that guidance from small-scale studies can be used with confidence to predict overtopping in prototype situations.

Current empirical methods often predict inadequately the overtopping discharges of waves on shallow sloping seawalls. As part of the DEFRA/EA funded research project FD2410, Coastal Flooding Hazard by Wave Overtopping, physical model studies have been undertaken for structure configurations of 1:2,1:10 and 1:15. For 1:10 and 1:15 sloping structures, no data or method is currently available for predicting accurately the overtopping discharge. Although discharge quantities are widely used for analysing the hazard of overtopping, this may not be the best way of defining a violent overtopping event. Discharge velocities may also be of key importance and it is these velocities that are reviewed here for overtopping of waves on sloping structures. Two numerical non-linear shallow water models have also been used for comparison against the physical results to see whether they can be implemented as a valid prediction tool.

A scale model of the NW Zeebrugge breakwater was tested under windy and windless conditions in the wind and wave tunnel of the Universidad Politécnica de Valencia (UPV) to study the effects of wind on runup and overtopping. The wind field is defined by a Pitot static survey within the wind and wave tunnel; runup is evaluated by a digital stepgauge, and the volume of overtopping discharge is driven into a tank and measured by a scale. As a result of the tests performed in the UPV, runup and overtopping were found to be increased by the wind. Additionally, the mean water level (MWL) is increased by the effect of wind while wind waves are produced in the flume riding over the mechanically generated ones. Windless experiments in the UPV and prototype measurements have been compared as available under the European Union (EU) funded project OPTICREST.

A modern high resolution numerical solver for the 2D shallow water equations on a Cartesian cut cell mesh is presented. The scheme is used to numerically model a wave basin containing a multi-element wave maker, two wave guides and a cranked seawall with a sloping beach. The model is run under a wide variety of wave conditions and basin geometries in order to provide guidance for setting up physical experiments within the wave basin to investigate violent wave overtopping of seawalls. Advice is provided on optimal wave guide length, the effect of separation between wave-maker and wave guides and the presence of overtopping hotspots at the seawall. A brief description of the Cartesian cut cell method is also given and this technology is used to simulate the multi-element wave maker.